Author: Raphaël Rocher

AI adoption operational readiness is usually an afterthought, and it always has been. Every technology a species has ever adopted arrived the same way: a capability first, a readiness question second, and usually not asked until something had already gone wrong. Fire didn’t wait for anyone to work out how to contain it. The printing press didn’t pause for someone to decide which lies were worth stopping. What’s different now is the speed of the gap between the two. A business can add a capability to how it talks to customers this afternoon and not discover what it actually needed to be ready for it until the first customer who mattered hits the edge of what it knows.

Right now, a lot of service businesses are being told the same thing: get an AI system talking to your customers before your competitors do, or get left behind. The question almost nobody in that conversation is asking out loud is whether the business is actually ready to be represented by something that doesn’t know what it doesn’t know.

The knowledge a call needs

Start with what makes a phone call to a business valuable at all. It isn’t the fact that someone answers. It’s that the person answering has, over years, picked up the parts of the job that never made it into a manual: the customer who called at 6pm on a Friday needing an answer the standard script doesn’t cover, the edge case that used to get walked straight to a manager because nobody junior had seen it before. That knowledge is expensive to build and easy to underestimate, because most of it was never written down. It lived in the person who handled the call.

Handing a customer conversation to an AI system assumes that knowledge already exists somewhere accessible, usually a website or a folder of internal documents. It rarely does, not in the form that matters. A website describes what a business wants a customer to know before they call. A phone call is where a business finds out what the customer didn’t already know to ask, and needed a person’s judgment to answer. Pointing a model at the first thing and expecting it to handle the second is a category error, not a rounding error.

The cost of getting this wrong isn’t abstract. In February 2024, a grieving grandchild named Jake Moffatt asked Air Canada’s website chatbot whether he could apply for the airline’s bereavement fare after booking his flight rather than before. The chatbot told him yes. It was wrong: Air Canada’s actual policy required the discount to be requested before travel. When Moffatt tried to claim it afterward, the airline refused, and then argued in front of the British Columbia Civil Resolution Tribunal that its own chatbot was, in effect, a separate legal entity, not something the airline could be held responsible for. The tribunal member called that “a remarkable submission” and ruled against the airline anyway: the company owns what it said, no matter what answered the question on its own website.

That is what happens when an AI system runs out of real knowledge and nothing stops it. Faced with a question outside what they know, a person in the same position does something an AI system doesn’t reliably do on its own: they say so, they escalate, they come back with an answer instead of inventing one. Air Canada’s chatbot didn’t run out of things to say. It just kept talking, with the same confident tone whether it was right or wrong, because nothing in how it was built distinguished between the two.

What escalation requires

It’s tempting to treat that as solvable with one more feature: build an escalation path, have the AI flag anything it can’t handle and route it to a person by email. That sounds like the fix. It’s only half of one.

A person who hits the edge of what they know and reports it to someone senior gets something an email flag doesn’t reliably produce: a correction that sticks. They get coached, they carry the answer forward, and the next customer who asks the same question gets a better response because a specific person learned something specific. An email escalation can technically do the same job, but only if someone reads it, understands it, and acts on it before the moment has passed. Anyone running a business today already knows how that usually goes when there’s a backlog and a dozen other things demanding attention in the same hour. The flagged email sits in a queue with everything else competing for the same limited attention a busy operator already doesn’t have enough of. The customer waiting on the other end of it has, in practice, been quietly deprioritized rather than helped.

AI systems can certainly be built to escalate well. Escalating well, though, is itself a piece of real operational work, not a checkbox a vendor ships by default. A business that hasn’t worked out how its own people escalate an edge case they can’t handle has no functioning template to hand an AI system in the first place. The tool can only inherit whatever discipline already exists, or the absence of one.

The wrong question people are asking

Not long ago, a public exchange between a software developer and a customer of a SaaS product showed how widespread this mistake is. The customer’s argument ran roughly like this: now that AI lets companies build software faster and cheaper, shouldn’t the price of that software come down? The developer’s answer was sharper than the question expected. What makes software expensive to run, he said, was rarely the writing of the code. It was everything downstream of it: the support tickets, the onboarding, the edge cases. All of that depends on operational machinery that has to work correctly around the software for the software itself to be worth paying for. Faster code doesn’t shrink any of that.

That exchange is a small, specific version of a much bigger mistake, and it’s the same mistake most businesses make when they reach for an AI phone system, an AI chat widget, or an AI-run intake process. It’s a close cousin of a mistake this site has already named on the marketing side, where selling a channel isn’t the same as having a strategy: the tool changes, the missing step underneath it doesn’t. The assumption underneath the purchase is that the problem was technological: if only responses were faster, if only someone always picked up, the business would convert more of what comes through the door. A tool magnifies whatever it’s handed, not what its buyer hoped for. Applied to a business that already runs cleanly, automation makes the clean thing faster. An AI system bolted onto a business that has never worked out who should be handling which decision, or why a call goes unanswered in the first place, just runs the same unexamined process faster, and runs the failures inside that process faster too.

Frederick Brooks made a related argument about software itself in a 1986 essay still taught in computer science courses forty years later: “No Silver Bullet.” Brooks split the difficulty of building software into two kinds: essential difficulty, the complexity genuinely inherent to the problem being solved, and accidental difficulty, the friction a team adds on top of that through bad tools or bad process. He argued that a single technology can only clear away accidental difficulty, never the essential kind. A business’s operational mess is essential difficulty: the decision flow nobody has ever mapped, the person handling three roles because nobody separated them. An AI system is very good at clearing accidental friction, and fundamentally unable to think its way through a business that never worked out what its own process was supposed to be.

What happens with no guardrail at all

The Air Canada case is what happens when an AI system runs out of knowledge on a real customer’s actual question. A second failure mode is what happens when there’s no guardrail stopping it from being pushed somewhere it was never meant to go, and the two failure modes look similar from the outside and come from the same root cause: nobody defined what the system should do when it left the path it was built for.

In January 2024, a customer of the UK delivery firm DPD deliberately provoked its chatbot until it swore at him and wrote a poem calling its own employer “the worst delivery firm in the world.” A few weeks earlier, a customer talked a Chevrolet dealership’s chatbot into agreeing to sell an $81,000 Tahoe for one dollar, adding “that’s a legally binding offer, no takesies backsies” for good measure. Both were manipulated on purpose, not confused by an ordinary customer, and that distinction matters: it means the failure wasn’t bad luck. It was the absence of any boundary on what the system could be talked into saying. A business that hasn’t decided what its AI system is not allowed to do has, by default, decided that anything is on the table if someone pushes hard enough.

Why this matters more here than most places

This industry has less margin for this kind of mistake than most. Movaros has written before about how thin the trade coverage is for moving companies, and how much of what an operator learns about running the business comes from word of mouth rather than any serious shared body of knowledge. That isn’t a criticism of any individual operator. It’s a structural fact about an industry that has never had the kind of information infrastructure other trades built decades ago. It also means the businesses in this industry are, on average, less prepared than most to catch an AI system’s mistake before a customer feels it. The operational discipline that would catch it is exactly the kind of infrastructure this industry has historically had the least of: documented decision-making, a real escalation path, someone whose job is to notice when something’s gone wrong.

Adding a capability on top of an operation that was never built to catch its own failures compounds that gap, at the exact moment a real customer, moving their whole life, finds out the hard way.

AI adoption operational readiness is the question worth asking instead

Here is where Harari’s own observation about technology cuts the other way, and it’s worth sitting with rather than rushing past. Every one of the tools people have adopted throughout history that changed how work got done did so by removing a specific, well-understood bottleneck, not by being generally impressive. The printing press didn’t succeed because it was clever. It succeeded because it solved an exact, nameable problem: the cost of copying a book by hand. The businesses that get real value from AI right now are doing the same thing at a much smaller scale: they find the one specific place in their own operation where a machine can genuinely do something a person was doing badly or slowly, instead of reaching for the newest capability and hoping a use for it appears.

That’s the actual first-principles question, and it has to be asked at the level of the specific decision, not the department. Not “should we use AI in customer service,” which is really no question at all. Instead: walk the actual flow a customer goes through, step by step, from call to quote to booking to move day. At every single step, ask: does a person need to be making this decision, or is this a place where the decision is mechanical enough that a machine handling it frees up a person for the parts of the job a machine genuinely can’t do? Some of those steps will have a real answer. A machine reading a Business Profile correctly is a straightforwardly mechanical task worth automating, the reason the right business even gets found in the first place. A machine improvising an answer to a grieving customer’s specific, unusual question is not, and pretending otherwise is how a company ends up explaining itself to a tribunal.

Hamilton Helmer’s own habit is useful here: ask which part of the process actually compounds in value over time, and protect that part specifically, rather than optimizing everything evenly. The knowledge a good employee builds handling real edge cases compounds. It gets better the longer that person does the job. An AI system with no mechanism for learning from its own mistakes just repeats whatever it was given, faster, whether that’s a well-built process or an unexamined one, compounding nothing.

Most of what’s being sold as AI readiness right now is really just AI availability. The tool exists, so the assumption is that using it counts as being ready. That assumption is the mistake. Readiness is the unglamorous, unfinished list of internal questions that were true before any AI system existed and are still true now: who actually makes this decision today, does that person have what they need to make it well, and what happens the moment they hit the edge of what they know. Solve that first, at the level of the actual decision, and an AI system becomes a genuine multiplier on a business that already works. Skip it, and the tool just makes the gaps that were already there move faster than anyone in the business can catch them.

If you handed your worst edge-case call from last month, the one that needed real judgment, to whatever AI tool you’re considering, would it have solved it, escalated it well, or made something up? That single question is a working AI adoption operational readiness test, and most businesses currently adopting these tools have never actually run it.

An agency told us we need retargeting, or social media, or a community. Is any of that actually a strategy? For most small logistics, relocation, and pet-transport operators who get sold one of these things, the honest answer shows up about a year later, buried in the invoices, and it’s no. What arrived was a channel: a tool that does one job, well or badly, for a monthly fee. What never arrived was the earlier, harder work of finding out why a specific customer, in a specific frame of mind, decides to hire this business at all. Skip that work. The channel ends up doing all the deciding. Whatever plan exists gets written after the purchase, not before it.

This isn’t a story about one dishonest agency. Most agencies selling this pattern to small, operationally-focused service businesses aren’t lying to anyone. They sell what they know how to sell, and what they know how to sell is a channel: retargeting, a stronger social presence, “building a community” around the brand. The failure underneath all three is identical and rarely named out loud: nobody did the granular work of understanding why this particular customer, feeling this particular way, actually decides. Operators are frequently just as guilty on their own side of the desk: they run one undifferentiated sales script across customers who are, psychologically, nothing alike. Assigning the blame to one side alone would be its own kind of dodge. The step that keeps getting skipped happens on both sides of the invoice.

The channel got sold. The strategy never showed up.

Here is the assumption nobody examines: that a small operator “needs to be doing digital marketing” in some general sense, and that any specific tactic offered under that banner is therefore a reasonable purchase. That assumption isn’t wrong so much as it’s the wrong altitude. Of course a moving company or a pet-transport business needs to reach customers. That’s not a strategic question. It never was. The strategic question sits one level down, and it keeps getting skipped: which customer, deciding under which conditions, is this specific tactic actually built to reach?

Retargeting works on someone who already visited a site and needs a nudge back. A stronger social presence works on someone who evaluates a business socially, through what other people say and show, before ever picking up the phone. “Building a community” works, if it works at all, on a customer relationship that outlasts a single transaction, the kind a gym or a subscription box earns, not the kind a household move typically is. None of these are bad tools. Each is built for one kind of decision, made by one kind of customer, at one particular point in that customer’s own reasoning. Sold without that match, a tool doesn’t fail loudly. It just runs, month after month, and produces exactly the kind of thin, hard-to-explain results that make an operator wonder if marketing simply doesn’t work for a business like theirs. Marketing worked. It was aimed at nobody in particular.

This site has already traced a visible version of the same failure from the seller’s side: an agency running an identical playbook across every operator it serves in the same metro, unable to produce a real edge for any one of them, because a template applied identically to different customers was never built for any of them specifically.

A related but different argument has already run on this site about what happens once an agency engagement ends: who keeps the rankings, the reviews, and the customer accounts an agency builds, a question of ownership. This is an earlier, more basic failure. Before anyone argues about who owns what a campaign produced, a prior question needs asking first: was the campaign ever built around understanding the customer it was supposed to reach, or was it built around a channel that happened to be for sale that quarter?

What Porter actually meant by the word strategy

The word “strategy” gets used constantly in small-business marketing conversations and defined almost never. Michael Porter’s 1996 Harvard Business Review piece, “What Is Strategy?”, remains the sharpest corrective available, precisely because it refuses to let the word mean whatever a seller wants it to mean that day. Porter’s central claim: strategy lives in a deliberately chosen set of activities, built around one clear customer position, not in the mere presence of activity. Buy a channel because it’s popular, or because a salesperson made a confident case for it. The result is something else entirely. In Porter’s own words, “a strategy is nothing more than a marketing slogan that will not withstand competition” once it stops being a genuine choice about where to compete and how.

The part of Porter’s argument that gets skipped most often in small-business marketing is the harder half: “the essence of strategy is choosing what not to do.” A retargeting budget, a content calendar, a community group: each one is a decision to spend limited time and money on that specific thing instead of something else. An agency pitching a channel rarely frames the sale that way, as a trade-off against every alternative use of the same budget. It gets framed as addition: one more thing a serious business “should be doing.” Strategy, properly understood, is subtraction first. Decide who the customer actually is and how they decide. Most of the available tactics rule themselves out immediately, long before any of them get bought.

Porter had a real precursor, and naming him matters because this mistake isn’t new. Theodore Levitt’s 1960 Harvard Business Review essay, “Marketing Myopia”, argued that industries decline not because their market runs out, but because leadership defines the business around a product or a channel instead of the customer need underneath it. A business that thinks of itself as “in the moving business” behaves differently from one that thinks of itself as making an anxious, high-stakes life transition feel manageable. Swap “moving business” for “a business that does retargeting.” The same sixty-five-year-old mistake reappears in a newer costume. Hamilton Helmer’s own strategy work exists for a related reason: he treats a word like “power” (the specific, durable barriers that let a company earn persistently higher returns than its rivals) with the same discipline Porter and Levitt bring to “strategy,” never a vague synonym for “being good at business.” Most operators buying a channel aren’t failing to work hard. They’re using a serious word loosely, and a loosely used word can’t tell anyone what to do.

Real rigor looks like Six Sigma, not a sales deck

If precise definition is the first missing piece, disciplined diagnosis is the second, and it already exists, fully worked out, outside of marketing entirely. Six Sigma began at Motorola in the 1980s as a rigorous approach to eliminating defects, and General Electric under Jack Welch turned it into one of the most consequential management disciplines of the following decade, according to Harvard Business School’s own Working Knowledge. Its core method, DMAIC, runs in a fixed order, as ASQ, the American Society for Quality, defines it: Define the problem, Measure it honestly, Analyze the real cause, Improve based on that analysis, then Control the fix. Nobody using DMAIC properly starts at Improve. Buying a channel before understanding the customer is exactly that: starting at the fourth step and skipping the first three.

Toyota’s own version of the same discipline is older and, for a small operator, more usable. Taiichi Ohno, the architect of the Toyota Production System, built the 5 Whys directly into how the company diagnosed problems on the factory floor, documented in his own 1988 book: ask “why” repeatedly, following each answer back to the next cause, until what’s actually broken becomes visible rather than whatever explanation happened to be offered first. Run it here. Why do we need retargeting? Because our website traffic isn’t converting. Why isn’t it converting? Because most visitors leave without booking a call. Why do they leave? Nobody in the business has actually asked. Three questions in, and the honest answer has stopped being about traffic at all. It has become a question about what a specific visitor needs to see or hear before trusting this specific business with a specific, high-stakes job. Retargeting might still be part of the eventual answer. It stops being the first thing bought, and starts being one possible tool chosen after the real question gets asked.

Neither DMAIC nor the 5 Whys is complicated. Both are available to any operator willing to sit with an uncomfortable question for longer than a sales call allows. That’s the entire, unglamorous difference between having a strategy and having a channel: one starts by defining the actual problem and asking why until the real cause is visible; the other starts by buying a tool and hoping.

The ASBFEO number that turns one frustration into a pattern

The pattern doesn’t stay theoretical. In January 2023, researchers at the University of the Sunshine Coast, working with the Australian Small Business and Family Enterprise Ombudsman, published the first study of its kind anywhere, nationally or globally, into exactly this relationship: small business owners and the digital marketing providers they hire. The study is Australian, worth saying plainly rather than burying, and the mechanism it documents has no obvious reason to stop at a national border. It surveyed 412 small business owners, backed by interviews with owners and providers on both sides of the relationship, and the pattern it found matches everything above almost exactly.

Around half of the owners surveyed said the recommendations their provider pushed on them were accurate, relevant, and worth the money. The other half said their provider aggressively pushed services that were too expensive or simply irrelevant to what the business actually needed. One respondent, in the hospitality sector, put it plainly: the services recommended “were not relevant to our business or not relevant to what we were wanting to do so, it was just kind of a waste of time and a waste of energy.” Roughly 30 percent of everyone surveyed said they had ended up in a formal dispute process over their digital marketing spend. The Ombudsman’s own office, separately, has actively reviewed 96 cases relating to digital marketing services since 2020, close to four a month, and receives a further dozen information requests through its call centre every month on top of that.

The same report has a second number worth sitting with, because it shows the mismatch isn’t a one-time bad hire. Nearly three-quarters of the owners surveyed, 73.5 percent, did not stay with a single provider longer than 12 months, and more than two-thirds had already used two or more providers by the time they answered the survey. That’s not a business slowly finding the right fit through trial and error. That’s the same purchase being made again and again, on the same untested assumption, with a different name on the invoice each time. A channel that was never chosen around a specific customer doesn’t get fixed by switching to a different channel chosen the same way.

Read those numbers next to Nassim Taleb’s recurring argument about who bears a risk and who doesn’t. The pattern stops looking like bad luck and starts looking like an incentive problem with a name. A provider’s fee isn’t contingent on whether the channel sold actually fit the customer buying it. The provider gets paid for delivering the channel, not for whether it was ever the right call. Every bit of the downside lands on the owner who wrote the check: the wasted budget, and in a real minority of cases, the formal dispute itself. A relationship that has run this way for years without visible failure isn’t evidence the model works. It’s evidence nobody with real skin in the outcome has stress-tested it yet. The 30 percent who ended up in a dispute are the closest thing this data has to that test, and the result wasn’t reassuring.

Operators run the identical mistake internally

The agency doesn’t carry this failure alone, and letting the blame land there would be its own kind of dishonesty. The same failure that shows up in a channel sold without understanding the customer shows up, just as often, inside an operator’s own sales process. This one wears the operator’s own handwriting instead of an agency’s letterhead: one intake script, one qualifying-questions sequence, one pricing conversation, run identically across customers who are, underneath the surface, nothing alike.

Apply the 5 Whys to that exact pattern too. Why do we ask every customer the same five questions on the first call? Because that’s just good, thorough intake. Why does thorough intake look identical for every customer? Because nobody has ever actually tested whether it does. Two questions in, and the real answer already has nothing to do with thoroughness. It’s an untested assumption, inherited from whoever wrote the script first, that every customer experiences the same five questions the same way. An operator who has never lost a sale to that exact mismatch has no evidence the script works. They have evidence it hasn’t been tested yet, which is a different claim entirely, and a much less comfortable one.

Picture two calls, back to back, on the same afternoon. The first caller has moved four times before, knows exactly what a binding estimate should include, and wants the call to move fast. The second has never done this, doesn’t know what to ask, and is listening as much for a tone that sounds trustworthy as for any specific answer. A script built around one set of default questions, asked in one default order, serves at most one of these two people well. It probably doesn’t fully serve either, because it was never built around a customer at all. It was built around a call. The fix costs nothing to test and nothing to buy: notice, within the first minute, which of these two people is on the line, and let that answer decide what happens next, not the script written for neither of them.

The tempting fix, on either side of this problem, is a bigger, more expensive version of the same mistake: a fancier channel, a longer script, a costlier version of the tool that never fit. The real fix is smaller and considerably harder: sit with the actual customer and decision long enough to know, specifically, why they say yes, before spending another dollar or another minute of a call on anything else.

The test that separates a strategy from a purchase

A channel is a tool. It is not a plan. Real strategy work looks like Six Sigma or the 5 Whys: define the actual customer and decision first, then choose the tactic that fits, and say no to the ones that don’t, the same trade-off Porter called the essence of the word. Most of what gets sold to small logistics and relocation businesses skips straight to the tool, and most operators buy it the same way, without ever forcing the earlier, harder question into the room.

The business consequence is specific, not abstract: money spent on a channel chosen before the customer was understood is money spent on a guess dressed up as a plan. It shows up as flat traffic that never converts, a social page nobody engages with, a “community” that never gets built. For close to a third of the owners surveyed in Australia, it shows up as a dispute over money already spent trying to fix all of it after the fact.

A direct test is available to any operator wondering which side of this they’re on, and it doesn’t require reading Porter or running a formal Six Sigma project. Ask an agency, in one sentence, why this specific customer responds to the specific channel they were sold, and don’t let the answer name the channel again as its own justification. If a straight answer isn’t available without circling back to the tool itself, the channel wasn’t chosen. It was offered, and bought, because buying something felt like progress. That’s not a strategy. It’s a purchase that resembles one, right up until the invoices are added up and somebody finally asks why.

Movaros exists on the other side of that same distinction: not another channel competing for the same budget, but the demand infrastructure an operator can build once the harder, earlier question about the actual customer has already been answered.

Two weeks ago this site argued that a real name and a real face beat a well-produced page, that the operators who reach the second, unaddressed decision-maker in a household are the ones who close deals an instrumental pitch alone can’t. That argument holds up. It was never meant to hold everywhere, and treating it as a universal law of trust would be a worse mistake than the gap it originally corrected.

Here is the question worth answering directly, because more than one operator has asked it since: we were told putting a real face and name on our sales process builds trust, so why doesn’t it always work? The honest answer is not a retreat from the earlier piece. It is a boundary on it, stated plainly now instead of discovered the hard way later. Relatability is a bet on the kind of purchase being made. Not every purchase is that kind, and the ones that aren’t punish the exact move that wins the ones that are.

Why putting a real face on it works in one sale and backfires in another

A full-service relocation is slow by design. Contact stretches over weeks, sometimes months, across multiple calls, and a household of two people forms an opinion of the company before a single box gets packed. Trust, in that sale, gets built the way any relationship gets built: repeated contact, a consistent name, evidence that one accountable person stands behind what happens on moving day. A photograph of the crew lead and a personal note in the quote email work because the customer has time to let that evidence accumulate, and because the thing being sold is inseparable from the person delivering it.

Pet transport runs on a different clock entirely. A customer calling about a flight-restricted breed or a last-minute international move isn’t shopping for a relationship. Most are already anxious. Some have already been turned away once by an airline that wouldn’t carry their animal. What they want from the first call is a functional answer: can you do this, what does it cost, how fast can you start. That customer evaluates the interaction closer to how they’d evaluate a flight booking than a real estate agent. Warmth offered at that moment doesn’t read as reassurance. It reads as a delay between the question and the answer, and delay is the one thing this customer has already had too much of.

Picture two calls landing on the same desk within the same hour. One is a couple relocating for a new job, six weeks out, still weighing three quotes at their own pace. The other is a family whose airline just told them, two days before departure, that their dog can’t fly cargo in this heat. Run the identical script on both: a warm greeting, a short story about the company’s history, a promise to treat their situation like family. The relocating couple hears care, and it moves them a little closer to yes. The family with a flight in forty-eight hours hears a company that hasn’t understood the actual size of their problem, and every extra second before the price and the timeline costs trust instead of building it.

Drawing on more than 75,000 real customer interactions, Harvard Business Review’s own research on this found that going out of the way to delight a customer has a negligible effect on loyalty compared with simply solving the problem fast. Most customers, in the researchers’ own words, just want a simple, quick solution. A pet-transport customer two days from a flight is that customer in its purest form, and a sales process built to delight is a sales process built for the wrong half of this industry’s buyers.

The research says the payoff from personal warmth depends on what’s being sold

This category gap is not a hunch dressed up as a rule. Robert Palmatier and his co-authors ran a meta-analysis of the actual research literature on relationship marketing, pooling results across dozens of independent studies rather than one company’s anecdote, and published it as “Factors Influencing the Effectiveness of Relationship Marketing” in the Journal of Marketing in 2006. Their headline finding is not that relationship investment works or doesn’t, but that how well it works depends heavily on the category being sold into: relational tactics pay off far more where the relationship itself is close to being the product, and far less where the customer is really just buying a defined outcome. Put a number on the gap: in categories where relationship building carries the most weight, the correlation with seller performance runs close to .58; in categories where it carries the least, closer to .43. That is a fifteen-point swing in how much the identical tactic is worth, decided entirely by category before a single word of the pitch gets written, not a rounding difference.

A relocation company sits close to the high end of that range, because the sale is a relationship in the fullest sense: a stranger is coming into a home, handling possessions with sentimental value, across a process that unfolds over weeks with room for a real bond to form. A pet-transport booking sits close to the low end, because however much the animal matters to the family, and it matters enormously, the actual transaction the customer is buying is a logistics outcome: safe, on time, done, with as little uncertainty in between as possible. The mistake was never building a relationship-first sales process. The mistake is assuming that process transfers unchanged from a category built for it into one that punishes it.

Here is the test worth applying before a single line of sales script gets written: is this customer buying an ongoing relationship with one person, or a bounded outcome from whoever is competent enough to deliver it. A relocation customer is buying the first, most of the time. A pet-transport customer, most of the time, is buying the second. Sell the first like the second and the pitch reads as cold. Sell the second like the first and the pitch reads as slow, and slow is the one thing an anxious customer will not forgive.

Warmth raises the stakes when something can still go wrong

A sharper reason than pacing explains why personal warmth backfires specifically in a high-anxiety, high-failure-risk category, and it has nothing to do with how long the sale takes to close. In the paper that defined how people judge others along two separate axes at once, Susan Fiske and her co-authors found that warmth and competence trade off in social judgment: the people and groups rated highest on warmth often get rated lower on competence, and the reverse holds just as reliably. Lead with warmth in a transaction where competence is the entire point, where the customer’s real question is whether this company can pull off something difficult and time-critical. There is a real risk that the warmth itself reads as evidence the company is soft on the part that matters most.

Marina Puzakova, Hyokjin Kwak, and Joseph Rocereto took that idea further and tested what happens specifically when a humanized brand fails at something. Their 2013 Journal of Marketing study, “When Humanizing Brands Goes Wrong,” found that giving a brand human qualities raises how much moral responsibility customers assign it. That sounds like a fine trade right up until the brand actually fails at something. A humanized brand that lets a customer down draws sharper, more personal blame for the identical failure than an unhumanized one draws for the same mistake. The warmth that made the company feel like a person is the same warmth that lets a customer treat its failure like a person’s betrayal, not a shipping error.

That is the real mechanism behind why a real name and a real face raises the stakes in pet transport specifically. It is not that pet owners are colder than home buyers. If anything the opposite is true: for a meaningful share of owners, attachment to a pet runs deep enough that a bad outcome is comparable in severity to losing a family member. It is that the category is already loaded with anxiety before the sales call even starts (a live animal, a tight window, a real chance something goes wrong that is outside anyone’s control). Putting a memorable person’s name on the process does not just build trust in advance. It builds a target for blame if that trust turns out to be misplaced. Customers criticize a generic, competent-sounding process as a process. They criticize a named, personally warm point of contact as if that person personally let them down.

Picture the shipment that goes wrong anyway, because in this category something eventually does. A connecting flight gets cancelled, a kennel run gets delayed at customs, a crate arrives four hours later than promised. If the company that handled it presented itself as a competent, well-run operation with clear procedures, the owner is angry at a process that failed and calls to ask what happens next. If the company that handled it was Sarah, who personally promised to take care of their dog like family, the owner’s anger has a name attached to it, and the call that follows is a different, harder conversation to have, for Sarah and for the business standing behind her.

The honest complication, and why the claim still holds

A fair reader might point to a different, better-known study here and ask whether any of this holds up against it. Xueming Luo and his co-authors ran a field experiment on more than 6,200 real outbound sales calls, some run by chatbots, some by human agents. They found that disclosing an AI agent as non-human cut purchase rates by more than 79.7 percent. If customers punish a company that hands them off to a machine that severely, the argument for staying deliberately faceless looks dead on arrival before it starts. It should. That is not the argument this piece is making, and the finding deserves to be named directly rather than quietly worked around.

The finding, read precisely, says something narrower than “people don’t want human contact.” Undisclosed, the chatbots in that study closed deals at nearly the same rate as experienced human agents, and roughly four times the rate of inexperienced ones, because they were scripted well enough to sound competent and in control. What collapsed the moment disclosure happened was not warmth. It was the customer’s confidence that a capable presence was handling their request at all. Every call in that experiment, chatbot or human, was still a scripted outbound sales call with no name attached to a face and no ongoing relationship implied on either side. The study measures human presence against machine presence. It says nothing about whether a customer wants the human already on the line to feel like a distinct, memorable individual they would hold personally responsible if the job went wrong later, which is the actual variable this piece is built around.

The two findings do not contradict each other. They describe two different failure modes sitting on two different axes. Luo’s result says a customer wants to know a competent, accountable presence is on the other end of the call, human or convincingly close to it. Nothing here argues against that. A pet-transport sales process still needs a real, reachable person who can answer a direct question fast, not a bot pretending otherwise. The narrower claim, the one this piece actually stands behind, is that once a real person is confirmed to be there, making that person feel personally, memorably warm raises the blame surface if the outcome later goes wrong. That is a cost worth paying in categories where the sale is a relationship, and a cost worth avoiding in categories where it is an anxious bet on an outcome. Personal warmth is a stake placed on a specific kind of sale. It was never a universal discount on trust.

Not the same argument as who is in the room

This piece and the earlier one are doing different work. Here is precisely where they part ways, rather than letting both blur into one soft claim about being human. The earlier piece on winning half a household argues that most relocation decisions have two evaluators, not one. A pitch built entirely for the instrumental reader, the one weighing price and reviews and coverage, leaves the expressive reader, the second person in the household, unaddressed. Its fix is adding a real name and a real face to reach a person the pitch was never written for in the first place. That argument does not touch category at all. It is entirely about audience: who else is reading the pitch besides the person who requested it.

This piece is about category, not audience. It argues that even a pitch aimed at exactly the right person can still choose the wrong tactic. The same tactic that reaches an expressive evaluator in a relocation decision is the tactic that raises the blame surface with an anxious pet-transport customer buying a bounded outcome. Reaching the right person and reaching them with the right kind of trust signal are two separate problems with two separate fixes. Solve the first without solving the second. A pitch can be perfectly targeted at the right household and still be wrong for the category it is selling into. An operator running both kinds of work, as plenty on this network do, needs both fixes at once, applied to the right jobs and never swapped between them.

What a category-aware trust strategy looks like

Start with the test named earlier, because it does the actual sorting work before anything else gets written: is this customer buying a relationship with one person, or a bounded outcome from whoever is competent enough to deliver it. Everything downstream follows from the answer.

For a relationship-purchase category, the earlier piece’s advice stands without qualification. Use a real name, a real face, a personal note. Give the second decision-maker in the house one accountable individual they can actually picture showing up. For a transaction-purchase category like pet transport, lead with the opposite of cold and corporate, which is fast and certain, not warm and personal. Answer the yes, the price, and the timeline before the introduction. Put the process front and center rather than a personality: what happens in what order, what the customer can check on, what happens if a flight gets cancelled two hours before departure. Introduce the person handling the job once the functional questions are answered, not before, so that introduction reads as reassurance about a decision already made rather than the reason to make it in the first place.

From the opening sentence, a category-aware first call for a pet-transport lead sounds different from a relocation intake call. It leads with the constraint, not a greeting: confirming the breed, the route, and the deadline, then giving a direct answer on feasibility and price inside that same call instead of promising a quote later. The reassurance about who is handling the crate comes after that, as a closing detail earned by a fast, competent answer, not as the opening pitch trying to earn the sale before the customer’s real question has even been addressed.

Whether that call reaches an operator through their own brand or through work the network sends their way, nobody upstream can judge this on an operator’s behalf. The category comes first. The tone of voice comes second, and it should follow from the category being sold into, not from a house style copied wholesale out of whichever past sale happened to teach the company the most about building trust.

So why doesn’t putting a real face on the sales process always work? Because it was never a universal law about human contact in the first place. It is a bet, correctly priced for a relationship purchase and badly mispriced for an anxious, failure-averse transaction. The same warmth that builds trust in advance is exactly what turns an ordinary shipping delay into something that feels, to the customer, like betrayal by a person instead of a mistake by a process. The fix is knowing which sale is standing in front of you before deciding which one to be, every single time the phone rings, not choosing between being human and being efficient.

Every operator has lost a deal that made no sense to lose. The reviews were strong. The quote came in competitive. The call went well, the customer sounded ready, and then nothing. Or worse: the job went to a competitor who charged more. The easy read is price. The real read, more often than not, is that the person who went quiet was never actually sold. Somebody else in the house was, and nobody ever asked that person anything.

Here is the question worth sitting with if that pattern keeps repeating: we seem to be doing everything right and still losing deals we should win, so what are we missing? The honest answer is usually that “we” was never the whole audience. A move, like buying a house, is rarely a decision one person makes alone. It’s a decision two people make together, and they don’t always look at the same evidence to get there.

A quote built for one signer

Most moving and relocation intake forms carry the same handful of fields: name, phone, email, move date, origin, destination. Every one of those fields assumes a single human on the other end, because the CRM behind the form was built that way. The CRM was built that way because it had to store the lead as one row with one contact. Nobody sat down and decided a household only has one decision-maker. The tooling decided it by default, and the marketing that was built on top of the tooling inherited that assumption without anyone re-examining it.

That default would be harmless if it matched reality. It doesn’t. The industry’s whole trust stack exists to convince one anonymous visitor: reviews, star ratings, association badges, all written for a reader who has already been reduced to a single row in a database. Everything downstream gets built to persuade that same name: the follow-up email, the quote PDF, the sales script. Nobody wrote the second half of the pitch, because nobody’s system ever asked who else was in the room.

The same logic runs through the analytics stack sitting behind the form. A conversion is one event, tied to one visitor ID, attributed to one channel. Marketing budget gets allocated against that single-visitor model because it’s the only model the dashboard can report on. Ask most operators how many people looked at the site before a job got booked. The honest answer is nobody tracks it, because the tools were never built to ask.

The math says otherwise. In the National Association of Realtors’ own 2024 Profile of Home Buyers and Sellers, 62 percent of that year’s home buyers were married couples, not solo buyers deciding alone. A household relocation skews the same direction for the same reason a home purchase does: it usually means a household, not one signer choosing on everyone else’s behalf.

Two decision-makers, two currencies of trust

Jagdish Sheth’s 1974 theory of family buying decisions is where this argument actually starts, decades before “conversion rate” was a phrase anyone used. Sheth’s foundational claim is the one nearly every later study on the topic builds from. Big-ticket purchases get decided jointly far more often than marketers assume, the kind a family makes rarely and lives with for years. The two people involved specialize by the kind of evidence they weigh, not by which one holds more authority in the relationship.

Real-estate research made that specialization concrete. Deborah Levy and Christina Kwai-Choi Lee interviewed nine experienced real estate agents in Auckland for a 2000 conference paper and found the same split holding in a market where nobody involved had read Sheth’s paper. Husbands, in their sample, specialized in what the paper calls instrumental factors: location, resale value, financing, the kind of evidence a spreadsheet can hold. Wives specialized in expressive factors: how a room feels, whether the layout works for the family that has to live in it, whether the agent seems like someone worth trusting with a six-figure decision. In higher socio-economic households specifically, the study found the wife typically contacted the agent first and stayed the main point of contact through the search. Neither role is decorative. Both evaluate, each from different inputs, and a pitch built entirely around one kind of evidence only ever reaches one of the two people who have to say yes.

The split isn’t fixed for the whole process either. Levy and Lee’s interviews describe five stages a household moves through: recognizing the need, specifying what they want, searching for options, evaluating alternatives, and making the final choice. The instrumental partner tends to carry more weight early, setting the location and price range, and again at the very end, negotiating price and terms. The expressive partner tends to carry more weight in the middle: doing the actual legwork of contacting agents, inspecting properties, and forming the gut sense of who’s trustworthy. A pitch aimed only at the price and the final negotiation misses the exact stretch where the second evaluator is doing the most work and forming the opinion that decides everything after it.

This is where a stricter definition earns its keep. A real competitive position has to hold against a specific kind of scrutiny, not scrutiny in general. Reviews and price comparisons are a defensible position against the instrumental evaluator: they answer whether this will work, and whether it’s worth the money. They are not a defensible position against the expressive evaluator, because that person isn’t asking whether the number holds up. They’re asking whether they can trust this specific company, these specific people, with something that matters to them. A business proven only against instrumental scrutiny has answered one kind of doubt and left the second kind unaddressed.

Real, credentialed academic research looks specifically at what happens to conversion when a company’s marketing deliberately addresses both members of a couple instead of just one. It was published in a peer-reviewed advertising journal in 2025. The direction of that finding matches everything above. Until it clears one more independent verification pass, the four sources already cited here carry the actual argument, and that’s enough.

What the numbers say about who’s in the room

The instrumental evaluator is easy to design for, because the numbers are the whole pitch: price, timeline, insurance coverage, a star average. The expressive evaluator is harder to design for, because what convinces them looks less like data and more like evidence of an actual person on the other end.

NAR’s own research on what buyers say matters backs this up directly, and not by a marginal gap. Buyers rated personal calls from an agent at roughly 71 percent importance, against 29 percent for the agent simply having a website and 14 percent for a social media presence. Whatever the industry has spent building out its digital presence, buyers keep saying the same thing when asked directly: a real, reachable person beats a well-produced page. The same body of NAR research found that 77 percent of repeat buyers interview exactly one agent before hiring them, most often through a personal referral rather than a comparison search. That’s not a market shopping on price. That’s a market shortcutting straight to whoever a trusted person already vouched for.

None of that is an argument against having a clear price, a strong review count, or a fast quote turnaround. It’s an argument that those signals answer the instrumental half of the question and stop there. A 71-versus-29 gap is a household telling an industry, repeatedly, what earns its trust, while the industry keeps optimizing for the number that lost.

The referral figure matters as much as the importance rating does, because a referral means someone the household already trusts vouched for a specific person, not a specific price: the expressive work, done in advance. By the time that referred household reaches the intake form, the instrumental case barely needs making. The form built for one visitor never won that job. A person did, working entirely outside the funnel the analytics dashboard can see.

The evaluator your website was never written for

Nielsen Norman Group’s Aurora Harley ran real usability sessions asking people to evaluate live company websites, and one moment from that research describes the expressive evaluator almost exactly. A participant landed on HomeCleanz, a cleaning-service site that asked visitors to submit an inquiry rather than showing pricing up front. She rejected it within 35 seconds. Her own words, recorded in Harley’s 2016 research: “I would definitely not use HomeCleanz because they don’t state the rate here, they want us to actually write to them. So I feel they are not open enough.”

Read that complaint carefully. It isn’t about price. HomeCleanz might have been the cheapest option on the page. The complaint is about openness, about a company that made itself harder to read than it needed to be at the exact moment a stranger was deciding whether to trust it. That’s an expressive-evaluator objection wearing the clothes of a pricing complaint. It’s close to the same objection a spouse voices when a moving company’s entire homepage is a spec sheet: services offered, coverage limits, a request-quote button, and nowhere a single photograph of an actual person who’ll show up on moving day.

A company can pass every instrumental test on that page (accurate pricing, clear coverage, a fast form) and still fail this specific reader in well under a minute, for reasons that have nothing to do with any of it.

This is a cheap fix by industry standards, which makes skipping it inexcusable. A photograph and two honest sentences about who does the work cost nothing next to a rebrand or a paid-search budget. The reason most operators skip it anyway isn’t cost. It’s that nobody on the team was ever assigned to write for a reader who isn’t asking about price.

Not ownership, not badge legibility: a second reader

Two other pieces on this site sit close to this argument. Here’s exactly where they stop and this one starts. Your reviews belong to the platform is about who owns a review once it exists: the platform-lock-in problem of an operator building a marketplace’s trust story instead of their own. Your trust badges are for other movers, not your customers is about legibility to a single stranger: whether a badge or star count means anything to the person reading it, as opposed to another operator who understands exactly what the credential costs to earn.

Both of those arguments still assume one reader. This one starts from two. A business can fix both problems: own its reviews directly, and replace its badge wall with something a stranger can parse. It has still only solved the problem for whichever evaluator happens to be looking. If the fix speaks entirely in instrumental terms (accurate numbers, real testimonials, a clear price), it has closed the legibility gap for one reader and left the second exactly where they started. The mechanism here is specifically that most of these decisions have two evaluators, not one, reading for genuinely different things. Fixing ownership and fixing legibility both raise the ceiling. Neither one reaches the second person by itself.

What it costs to reach both, and what it’s worth

Picture a household requesting quotes for a full-service relocation this month. If NAR’s own numbers hold even loosely for a household move the way they do for a home purchase, on well over half of those calls a second decision-maker is going to open the same page later that night, after the person who made the call has already formed an opinion. If the page they land on repeats exactly what the first person already saw (the price, the star count, the coverage terms), it tells the second person nothing they didn’t already have secondhand. It answers a question nobody in the room was still asking.

Two operators are quoting the identical job, side by side. Operator A’s page carries forty reviews, a live quote calculator, and a coverage table. Operator B’s page carries twelve reviews, the same coverage table, and one addition: a ninety-second video of the actual crew lead introducing himself and walking through what moving day looks like. On the instrumental read alone, A wins comfortably: more reviews, a slicker tool, possibly a lower price. But when the second decision-maker, having never heard the original call, opens both tabs that evening, B answers a question A’s page never tried to: who is actually going to be in my house.

What changes that isn’t a redesign. It’s usually one addition: a real name and a real face attached to the estimate, a two-line note in the quote email from the actual person who’ll run the job, a phone number that reaches a human within the hour instead of a ticket queue. None of that costs what a new website costs. It costs someone deciding to write to the second reader on purpose, instead of hoping the first person’s enthusiasm carries the whole room. It is also the same content already winning the discovery race moving to social feeds, not just the trust race inside an inbox.

The obvious objection: what about the households where there really is only one decision-maker? Plenty of moves get booked by someone acting entirely alone, and nothing above argues otherwise. But a business built to reach exactly one has capped its own ceiling at whichever share of its market decides alone, with nothing at all built for a second reader. NAR’s 62 percent is not every household, but it is most of them, and “most” is too large a number to leave unaddressed by design rather than by choice.

Ask the other person who said no

Most operators who lose a deal that should have closed land on the same explanation: the price must have scared them off. That answer is comfortable because it doesn’t require rethinking anything. It’s also wrong more often than the industry wants to admit, because a lost sale with two decision-makers in the house only needs one of them to have a reason to say no, even while the other one was ready to sign.

This isn’t a call to redesign anything expensive. It’s a call to stop being surprised. An industry that keeps blaming price for a loss it never diagnosed would rather feel unlucky than look at its own funnel. The data above isn’t subtle: a household is usually two people. They weigh different evidence, and most businesses built their entire pitch for exactly one of them.

Marketplaces and badge walls aren’t the villain here, and nothing above claims they are. The real charge is narrower: a business that speaks fluently to one evaluator and not at all to the second has quietly capped its own close rate at whatever share of buyers happen to decide alone, then called the rest a pricing problem instead of what it is. Optimizing every signal for a single evaluator’s trust caps conversion at whichever half of the household that signal reaches. The other half’s no is enough to lose the sale, even when the first person was already fully convinced. The same logic favors building on a brand a household can actually meet over renting a listing that a marketplace controls entirely. An operator running its own name is the only one that can write to that second reader on purpose, because it’s the only one who knows a second reader exists.

If the two people who made an operator’s last lost decision each explained separately why they walked, would they give the same reason? For a lot of operators, the honest answer is that nobody ever asked the second person anything. That’s the actual gap. Not price, not competition, not a review count that came in slightly lower than a rival’s. A pitch that fully convinced exactly one person in a two-person decision, and never spoke to the other one at all.

Business is up. Real money got you there: a faster site, quicker follow-up, maybe a marketing agency finally worth the invoice. But someone still answers the phone, reads the enquiry first, and decides whether to chase a hesitant lead or let it go cold. That person sounds exactly like they did before any of that changed. Ask most owners why. The answer comes fast: I need better people, or I should pay them more. Both assume the problem lives inside the person answering. It doesn’t. It lives inside the deal struck with them, and that deal was never built to reward what growth now asks of it.

Here is the question underneath that one: you want to win more business, so why does it feel like nobody on your team is actually pulling in that direction?

Why nobody on your team pulls in the same direction you do

The people fielding customer contact at a moving, freight, or pet-transport business are almost never the owner. They’re salaried coordinators, hourly staff, sometimes cash-paid crews pulled in for a busy week. None of them personally profit from the gap between a booked job and a missed one. Paid twenty-two dollars an hour to answer intake calls, a coordinator captures none of the marginal revenue when a four-thousand-dollar move books instead of falling through. Whether they chase eight leads that day or eighteen, the paycheck reads identical on Friday. The same holds for a freight dispatcher deciding which lane to push harder on, or a pet-transport coordinator choosing whether to call a nervous client back today or tomorrow. None of them own equity in the business. None of them see a cent of what a stronger quarter is worth to the person who does. A system built to send that person more leads doesn’t raise their pay. It raises their workload, at the same rate, at the end of the same week.

This site has already measured what that produces. Why enquiries go quiet found that a majority of enquiries get a reply too late to matter, and traced it to slow response, missing after-hours coverage, and quotes nobody chases back up. That’s the symptom, measured correctly. It doesn’t ask why a business that genuinely wants the work keeps producing that exact behavior, month after month, regardless of who’s staffing the desk that week.

The honest answer isn’t a training problem. It’s an incentive problem. Once that’s visible, the slow reply stops looking like negligence and starts looking like the predictable output of the system that produced it.

Economists gave this exact gap a name in 1976

Michael Jensen and William Meckling’s 1976 paper in the Journal of Financial Economics, Theory of the Firm, is the founding document of what’s now called agency theory: the study of what happens whenever one person, the principal, hires another, the agent, to act on their behalf, and their interests don’t fully line up. The paper opens by borrowing a warning from Adam Smith: someone managing another person’s money will rarely watch it with the vigilance they’d apply to their own. Fifty years on, the warning describes exactly what happens at a phone desk. The owner is the principal. The person answering is the agent. The agent has no built-in reason to treat a stranger’s enquiry the way the owner would.

Jensen and Meckling also named the resulting waste: agency costs, the sum of what a business spends trying to keep an agent aligned. That spending shows up as monitoring, oversight, and tighter scripts. But agency costs also include what the agent still gives up to protect their own interests anyway, and the value that’s simply lost in between. Most owners are already paying the first kind, in call-monitoring software and mandatory check-ins, without touching the reason any of it is needed in the first place.

This isn’t theoretical. Steven Levitt and Chad Syverson put a number on it in 2008, using a market where the same person plays both roles at different times: real estate agents. Their study in the Review of Economics and Statistics found that agents sell their own homes for 3.7 percent more, and leave them on the market 9.5 days longer, than they achieve for an identical client’s house. Same person, same skill, same market knowledge. The only variable that changed was who kept the extra money. When it’s the agent’s own equity on the table, they hold out for a better number. When it’s a client’s, they take the first workable offer and move to the next file, because their commission barely moves either way and their own time does.

A phone desk runs the identical arithmetic. Nobody just writes it down. The person capturing that lead well chases it, works the follow-up, and gets the quote out fast, but gets none of the marginal revenue a closed job represents. They get the same paycheck whether the enquiry converts or dies quietly. What they do get, if they push harder and it works, is more of exactly this next week, at the same rate. A system that’s run smoothly for years on this basis looks stable only because nobody has asked it to do more yet. Stability that has never been tested by real growth isn’t evidence the incentives are working. It’s evidence the test hasn’t happened.

Does paying people more actually fix it?

The tempting fix is obvious: pay more, or add a bonus tied to bookings. Alfie Kohn’s 1993 Harvard Business Review piece, Why Incentive Plans Cannot Work, remains the sharpest institutional answer to that instinct, and it isn’t encouraging. Kohn’s review of decades of workplace and laboratory research found that rewards reliably buy one thing: short-term compliance on simple, easily measured tasks. On anything requiring judgment, persistence, or genuine care, a bonus tends to narrow effort down to exactly what gets measured, and nothing past it.

The counter-case is real too, and an honest piece has to sit with it rather than argue around it. Edward Lazear’s 2000 study in the American Economic Review, Performance Pay and Productivity, tracked roughly three thousand workers at Safelite Glass after the company switched windshield installers from an hourly wage to piece-rate pay. Output per worker rose 44 percent. Profits rose with it. Incentive pay, done right, genuinely works.

The two findings sit together once you look at what made Safelite’s bet succeed. Installing a windshield is a task with one countable output, attributable to one person, with quality easy enough to check on the spot. A phone desk is nothing like that. Persuading a hesitant caller, chasing a quote through three follow-ups, staying warm with someone who isn’t ready yet: none of it reduces to a single number one person can be paid against without inviting exactly the corner-cutting Kohn documented. Pay a flat bonus per booked job and watch coordinators start pressuring undecided callers instead of qualifying them properly. The fix that worked on a windshield doesn’t transfer cleanly to a conversation.

The skeptical owner’s real objection belongs here: I’ve hired great people before, and they still don’t chase hard enough. That’s likely true, and it isn’t evidence against any of this. A strong hire dropped into the same unrewarded structure behaves like everyone else in it within a few months, because the deal, not the person, sets the ceiling on effort. Swapping the person without touching the deal just resets the clock on the same outcome.

The research that does transfer is less comfortable than a bonus scheme. Harvard’s Heskett, Jones, Loveman, Sasser, and Schlesinger built the service-profit chain from studying companies where frontline performance drove profit, and the lever wasn’t sales pressure. It was internal service quality: the tools, the authority, and the working conditions given to the person doing the job. Those elements drove employee satisfaction, then retention, then the customer experience that shows up as revenue. Two things can be true at once, and the honest version of this argument holds both without picking a side: incentive pay can work, and most businesses reaching for it are applying a windshield-shop fix to a job that never resembled one.

The deal never gave the person answering the phone a reason to care whether the business wins more work. It gave them a reason to avoid taking on more of it for the same pay. That isn’t a motivation problem waiting on a pep talk or a better hire. It’s structural, built into the deal itself, and it explains why growth stalls even when an owner genuinely wants it and is doing everything else right.

What happens when nobody fixes it and the market does instead

Businesses that never correct this misalignment eventually meet a competitor who solved it structurally. That competitor built a model where the person closest to the customer carries real exposure to the outcome. Three cases, decades apart and nothing to do with logistics, show the identical mechanism at three different scales. These aren’t stories about a scrappy underdog outworking a comfortable giant. Each one is a story about a cost structure crossing a threshold: once the marginal dollar started flowing to whoever closed the sale, the businesses on the other side of that shift never had a real chance, no matter how much they spent trying to compete on service or marketing instead.

A hotel’s front desk clerk earns the same wage whether the room sells or sits empty. An Airbnb host earns nothing unless the listing books, and keeps the difference personally when it does. That single difference in who’s exposed to the outcome is a large part of why the hotel industry’s revenue moved when Airbnb entered a market. Georgios Zervas, Davide Proserpio, and John Byers, economists at Boston University, measured it directly: every 10 percent increase in Airbnb listings in a market cut hotel room revenue by roughly 0.39 percent on average. In Austin, Texas, where Airbnb’s presence grew fastest and deepest, the cut reached 8 to 10 percent. Hotels didn’t lose that revenue because their staff got worse at the job. They lost it because a host with direct equity in every booking will out-hustle a desk clerk with none, every time the two compete for the same guest.

The same mechanism decided a format war a generation earlier, with nothing to do with picture quality. Sony kept Betamax’s licensing closed, manufacturing it alone and controlling every detail of the format. JVC did the opposite with VHS, licensing it to Matsushita, Hitachi, Sharp, and a dozen other manufacturers, each one now personally invested in VHS’s success because their own sales depended on it. Michael Cusumano, Yiorgos Mylonadis, and Richard Rosenbloom traced the outcome in the Business History Review in 1992: VHS players got cheaper faster, more manufacturers meant more tape titles worth stocking, and more titles pulled in more buyers, a loop Betamax couldn’t match because only one company had anything riding on it. Sony’s own leadership later acknowledged, well after the war was decided, that keeping the format closed had cost them the market they were trying to protect. A coalition where more people carry a real stake in the outcome beats a monopoly on control more often than not, because more of the people who could make it succeed have a reason to.

The starkest version of this mechanism is what happened to New York City’s taxi medallions, and it’s the one that deserves the most care in how it’s told. This isn’t a clean story about an upstart beating a slow incumbent on efficiency. It’s a story about who was exposed to a bet and who wasn’t. The people who ended up carrying all of the loss were the ones with the least control over how the bet got made.

Between 2004 and 2014, the price of an individual taxi medallion at auction rose from $283,300 to $965,000, according to a 2020 lawsuit that New York Attorney General Letitia James filed against the city’s own Taxi and Limousine Commission, alleging the agency ran its auctions in a way that artificially inflated those prices for over a decade. A Democracy Now interview covering the crisis reported that the city collected more than $855 million from medallion sales across that period, and bore none of the risk if the value it had helped inflate ever came back down. Lenders wrote loans against those prices and collected origination fees and interest regardless of what happened next. Neither the city nor the lenders had a dollar of their own tied to what the medallion would be worth later.

The drivers who bought those medallions did. The same interview found that roughly 91 percent of New York’s taxi drivers were born outside the United States, many without fluent English, and that a number of them did not fully understand the terms of the loans they signed. When ride-hailing arrived and pulled fares away, medallion prices collapsed. By mid-2019, individual medallions were selling at auction for as little as $137,000 to $138,000, according to Crain’s New York Business, a fraction of what many owners had borrowed against them just a few years before. At least eight drivers, including three medallion owners, have died by suicide since 2018, and reporting on the crisis has directly connected several of those deaths to the debt.

This deserves to be said plainly: none of it should read as a tidy lesson about disruption rewarding the efficient. It’s a story about a bet that only ever had one set of people exposed to losing it, set by parties who had nothing of their own at stake and walked away regardless of how it ended. That’s the same mechanism as the phone desk and the hotel counter, at a scale that cost people their lives, and it deserves to be remembered as exactly that, not filed away as a punchy statistic about an app beating a taxi.

The alternative to waiting for a competitor to fix this for you

Two of those three examples share something worth naming plainly. The hotel clerk and the medallion driver were both incumbent workers, and both disruptions landed at their expense, not at the expense of the executives who ran the systems that failed them. That’s the version of this story most operators fear, and they’re right to. It doesn’t have to be the only one available. The partnership structure behind Movaros’s own fulfilment page exists for a different reason: a smaller operator plugs into a network with real demand behind it and grows because of the arrangement, instead of waiting to see which competitor solves this incentive problem first and takes the market while everyone else is still hoping a pep talk works.

Run this test before the next growth push

Every fix described so far corrects a structure, not a person. Before reaching for a new hire, a script, or a bonus plan, run a smaller test first. Picture the business landing 20 percent more qualified work next month than it has today. Now picture the specific person who answers the phone or opens the first enquiry each morning. Would that person be better off with a bigger share of the extra revenue, more authority to make a call without checking with someone else, a clearer stake in whether the extra work turns into extra pay? Or would they just be busier, doing the identical job, for the identical rate, with more calls left unanswered by closing time?

Most owners, asked directly, already know the honest answer. It explains far more about a slow reply, an unchased quote, or a caller who never gets a real answer than any story about a lazy hire ever will. The fix was never a better person. It’s building the one thing the phone desk has never had: a reason to want the growth as much as the owner does.

When we sat down and reviewed real, live moving-company sites for this piece, we did not go looking for the neglected ones. Several were the opposite, run by operators who clearly take the web seriously: current photography, genuine reviews, years of visible investment. That is why the pattern that follows is worth writing down, not filing away as one company’s oversight. It shows up on well-regarded, well-invested sites about as often as it shows up on tired ones. Almost nobody goes looking for it, and the people who built these pages were each optimizing for something reasonable that had nothing to do with the one thing the page is for. That pattern has a name: website conversion leaks. Aggregators charge for exactly this kind of unforced error, multiplied across an entire market.

A shop assistant greets every customer personally and answers every question. Then, the moment the customer reaches for their wallet, the assistant hands them a flyer for the store across the street. That assistant would not last the week. Most operators’ websites do a version of that more than once before a visitor ever reaches the quote form, and the operator paying to bring that visitor through the door has usually never noticed.

That is the real question underneath this piece: is your own website quietly sending the customers you already paid to reach somewhere else? Not through a defect a stranger built into it. Through choices that felt reasonable, even generous, at the moment someone made them.

The trade the aggregator is making

This site has already described twice what that trade costs when a marketplace makes it on purpose. The six-quote problem walks through what happens once an operator’s quote enters a platform’s blind auction: real qualification time spent, and a result with no reason attached. Every transaction also teaches the platform something it sells back to the next operator at a slightly higher price. The distribution tax puts a number on the other half of the same arrangement: the dollar cost, compounding over time, of renting demand from someone else’s gate instead of owning the road to your own door.

Both pieces describe a transaction. A platform takes a visitor who already decided to act, and in exchange for access to that visitor, it charges the operator, learns from the exchange, and keeps the relationship for itself. It is not a friendly arrangement, but it is at least an honest one. The platform is charging for something real, and everyone involved knows roughly what the price is.

Almost nobody has priced the version of that exact trade an operator’s own website runs on itself, for nothing.

Is your own website quietly sending customers somewhere else?

Ask most operators why their header carries three social icons, why the “read our reviews” link goes straight to an outside platform mid-quote, or why the navigation bar lists fourteen items. The honest answer is some version of: social links and a reviews page build trust, and a full menu means we’re not hiding anything. Every one of those instincts is reasonable on its own. Put together on the one page whose entire job is to keep a paid-for visitor there long enough to convert, they add up to a site handing its own traffic back to the open web, one click at a time, at the exact moment a visitor is deciding whether to stay.

Three specific exits turned up on page after page during this review, on sites that otherwise looked considered and current.

The header that offers six ways off itself before the pitch loads

Three, four, sometimes five icons sat across the top of the pages reviewed for this piece, each one a live link to Facebook, Instagram, YouTube, or LinkedIn. Each was weighted on the page exactly as heavily as the phone number and the quote button beside them. Every icon is a door. Click one and the visitor leaves the site, built and paid for to sell them a move. The visitor lands somewhere an algorithm decides what they see next, with no guarantee any of it points back.

A real, if narrower, body of research explains why crowding a decision point with more options costs something. Sheena Iyengar and Mark Lepper’s famous jam study deserves precision, because it gets misquoted constantly: shoppers offered a table of 6 jams bought at roughly ten times the rate of shoppers offered a table of 24. The study happened, and it found that. It does not prove that fewer options always convert better everywhere, and a decade of follow-up work has made that clear. Benjamin Scheibehenne and colleagues’ 2010 meta-analysis pooled 63 conditions and more than five thousand participants and found the average effect hovering near zero. Its confidence interval straddled no effect at all. A direct attempt to rerun the original jam experiment in an upscale German supermarket found nothing. Choice overload is real under some conditions and absent under others, and nobody has fully mapped which is which.

The header icons undermine something narrower and older than choice overload: attention on a single, distinctive next step. Hedwig von Restorff’s 1933 finding, still holding up in modern review, was about memory rather than clicks: an item that stands alone against a field of similar items gets noticed and recalled far better than one lost in a row of equals. A quote button surrounded by four social icons of matching size and color is not isolated. It is one option among five, on a page where only one of the five was ever supposed to matter.

The review link that hands the visitor back to the marketplace

The second exit is more specific, and more expensive. A “read more reviews” link sits right where a visitor is comparing this operator against two or three others, and it goes straight to that operator’s public profile on an outside review platform. That page is one the operator does not control. It surfaces the map, the competitors nearby, and often a panel of similar businesses, pointing at exactly the names the visitor was trying to choose between. The operator built a page to make its own case. The reviews link, at the precise moment of comparison, walks the visitor over to the one page on the internet built to show them everyone else’s case too.

A popular claim in UX writing holds that Amazon deliberately strips navigation out of its own checkout flow for exactly this reason, to remove every possible exit at the moment of decision. That claim deserves honesty: it does not trace to anything Amazon has published, patented, or said on the record. It is widely observed and repeated in UX circles, though Amazon has never stated it as deliberate policy, and it should carry no more weight in this argument than that.

What Amazon has actually said, on the record, carries the real weight here. An Amazon director who helped shape the product strategy behind Alexa and Amazon Music, Kintan Brahmbhatt, named the exact mechanism directly: “Context switching often happens when a customer must navigate away from your app or site to complete a task. It’s the point at which your customer will abandon your product.” That is not folklore. A named Amazon leader describes, in a credible operator publication, precisely what an off-site reviews link does mid-quote.

The mechanism has support outside Amazon too. Diana DeStefano and Jo-Anne LeFevre’s review of hypertext reading found that every link a reader encounters mid-task is a small decision of its own, weighing whether to follow it, and that decision load measurably drags on completion of the task the reader came to finish. A reviews link is not neutral information sitting quietly on the page. It is a decision the visitor has to make, at the worst possible moment to be handed one.

The natural objection is fair: don’t reviews build trust, and doesn’t hiding them look worse than showing them? The objection conflates two different things. Showing proof is not the problem. Sending someone away to go find it is. The star rating, the review count, three or four specific quotes with names attached, can all sit directly on the page, in the operator’s own layout. That does the identical trust-building work, without opening a door out of the funnel to do it. It’s the same principle a badge wall gets backwards: specific and attributed beats generic and borrowed, whether the proof is a review or a badge.

The menu built for the org chart, not the visitor

The third exit sits in plain sight on almost every page reviewed: a navigation bar carrying somewhere between ten and twenty items, organized the way the company organizes itself rather than the way a visitor thinks about their own move. Residential moving, commercial moving, storage, packing services, international, careers, blog, about us, FAQ, financing: each one is a tab, and each one competes for the same strip of space a visitor scans in the first few seconds on the page.

Baymard Institute’s research on product-catalog navigation, based on testing across 344 sites, found that 75 percent of them overcategorized their navigation in ways that measurably cost them customers. The reasons come down to two: visitors could not compare things across categories, or they wrongly concluded something was not offered at all because the category they expected did not contain it. A visitor scans a mover’s menu for a way to get a quote, past eleven other tabs, and runs the identical risk: wrongly concluding that the thing they came for is buried somewhere they have not found yet, and leaving rather than continuing to hunt.

It would be convenient to say the fix is a hard item limit, but Nielsen Norman Group’s own research explicitly rejects that. Kathryn Whitenton’s analysis of navigation menus found no support for a fixed rule like seven items or fewer. The real defect is not the count. It is whose logic built the list. A menu organized around what a business does internally (departments, history, press mentions) asks a visitor to translate their own task into the company’s language before they can act on it. A menu organized around what a visitor came to do puts the one thing that matters, requesting a quote, first and unmissable, and lets everything else recede.

What the echo costs

Every instinct behind these three choices sounds like good practice. Show your social proof. Be transparent. Do not hide your menu. Flip each one and the logic breaks down fast: a shop’s whole strategy is not usually to advertise the exits on the way to the register. The platforms this industry complains about the loudest spend enormous effort removing exactly this kind of friction from their own funnels, then build a business selling operators access to the visitors who stayed. An operator’s own website can copy the aggregator’s everything-available instinct. But it skips the part where the aggregator gets paid for the traffic it captures. That is not more honest. It is the same funnel, running at a loss.

None of this is a story about any one team being careless. It is structural, in the way an industry-wide pattern usually is: a handful of web-design agencies, review-widget plugins, and template libraries serve most of this industry, and a defect built into a template propagates to every business that licenses it, regardless of how much it cares. A well-run operator and an average one can end up with the identical header, the identical review widget, the identical menu logic, because neither one built the page from first principles. Every site carrying that template carries the same exposure, and no amount of individual diligence downstream fixes an assumption baked in upstream.

Why does a page built by people who genuinely care still leak visitors this way? The honest answer has little to do with competence, and everything to do with incentives. The agency that built the header gets judged on whether the client likes how the page looks, not on quote-form completions six months later. The review platform wants its badge clicked, because that click is how it proves its own value to the operator paying for it. Nobody in that chain was ever paid to protect the fifteen seconds between a visitor landing on the page and deciding whether to stay. The page ends up optimized for everyone’s goals except its own.

Borrow the model the distribution tax already built to put a number under this. A 100-move operator doing $500,000 a year and spending 10 percent of it on acquisition is paying $50,000 a year for a pipeline of visitors who are genuinely in the market to compare a move. That $50,000 buys attention. It does not, on its own, buy a conversion; the page still has to close what the marketing bought. Picture that budget landing a thousand qualified visitors a month, in round numbers. Walk a modest share of them through the three exits above: one in fifteen follows a header icon and does not return within the session, one in twenty clicks the reviews link into the marketplace’s own comparison page mid-decision, one in twenty-five gives up hunting through a menu that never gets them to a form at all. None of those figures is a verified industry statistic. Nobody has run a controlled study of this exact funnel on this exact industry’s sites, because until now nobody had reason to. But even a conservative version of that stack quietly voids a real slice of a $50,000 media budget before the page has done the one thing it was bought to do. The real number for your own site isn’t a guess. It’s sitting in the same two numbers that tell you whether a website is working at all.

The six-quote problem describes what a marketplace charges an operator for doing this on purpose, at scale, as a business model: taking a visitor who was ready to act and handing them somewhere else to go instead. A website that opens several of its own doors mid-decision is running the identical playbook against its own paid-for traffic, for free, with nobody on the other end even learning anything from it. The marketplace at least turns the transaction into data it can sell back to the next operator. A homepage that leaks the same way just loses the visitor and calls it normal.

The walk-through that costs nothing to run

The fix for all three exits needs no developer, no redesign, and no marketing budget. It needs someone at the business to do what a stranger comparing several movers under stress does: land on the homepage cold and click everything on it that is not the quote button.

Count the icons in the header first. If clicking one takes a visitor off the site with no path back into the quote flow, that icon is costing more than it proves. A social presence is worth having. It is not worth advertising at the same visual weight as the one action the page exists to produce. Most operators find they can drop two or three icons entirely and keep exactly the same social presence, just without a header full of live exits.

Follow the reviews link next, from a fresh, incognito tab, the way a visitor would mid-comparison. If it lands anywhere that also shows competitors, the fix is not to remove the proof. It is to bring the proof home: pull three or four specific, attributed reviews, quotes with names attached rather than a star average, directly onto the page, and let the outside platform stay a citation rather than a destination.

Then walk the navigation bar and ask, honestly, whether each item answers a question a visitor actually has, or a question the business wanted to answer about itself. Careers, press mentions, and company history are legitimate pages. They are not homepage-navigation-tier pages, competing for space with the one link that turns a visitor into a lead.

None of that fixes what an aggregator charges for access to a customer who was already looking to book. It fixes something cheaper and entirely within an operator’s own control: the version of that exact tax the business has been charging itself, for free, every time a visitor who was already there decided to leave.

Run the count for real. Open your own homepage in a fresh tab right now and click every link on it that leads off your own site during an active task, the way a visitor mid-decision would. However many you find, ask which ones were put there on purpose, and which ones just accumulated. That answer, not a redesign budget, is where this gets fixed.

A shop assistant who did that would get corrected by lunchtime, because someone would be standing there to see it happen. Nobody stands at a website’s shoulder watching every exit a visitor takes. That is exactly why it survives, on good sites and average ones alike, until an operator finally goes and looks. Website conversion leaks. Aggregators never have to go looking; they built a business on the fact that most operators don’t.

Forty years in business. Thousands of moves done properly. A crew that knows what they’re doing and a claims record most competitors would envy. None of it matters to a customer who never finds out the company exists.

That’s not a rhetorical flourish. It happens mechanically, every time a moving decision starts on a platform instead of with a name someone already trusts.

The path has changed, whether or not the reputation has

A customer planning an international move increasingly doesn’t start with a Google search for a specific company name. They start on a comparison site, or type a general query into an AI system, or land on a directory. Sirelo alone lists more than 26,000 movers and says it put over 200,000 consumers through a comparison process in 2025. Somewhere in that funnel, a decision gets made about which handful of operators the customer sees: comparison site, search engine, AI assistant, broker.

An operator’s history, accreditations and reviews only start to matter once they’re inside that shortlist. The decades of good work that happened before the shortlist are invisible to a process that was never built to weigh them.

Run the arithmetic on that pool. Assume, for illustration, that a typical comparison result surfaces somewhere between five and eight movers to any one customer; Sirelo doesn’t publish that figure, but it’s a reasonable range for how these shortlists tend to display. Out of 26,000 listed movers, the math puts any single operator’s odds of appearing in front of a given customer at roughly two or three in ten thousand, on a purely random draw, before service quality enters the picture at all. That’s the size of the gate a reputation now has to pass through before it can do any work.

The deeper shift is in what starts the search in the first place. In the old path, a customer already had a name before they searched: a neighbor’s recommendation, a name from a past move, a company seen on a truck in the right part of town. The search was for that name specifically. Brand reputation converted directly into being found, because the customer supplied the brand. In the new path, the search starts from a category, not a name: “movers from London to Sydney,” typed into a comparison site, a marketplace, or an AI assistant. Nobody supplies the name. Something else has to.

Reputation and visibility are not the same asset

It’s easy to assume they compound together, that being good at the work eventually shows up as being found for the work. For a long time, in a lot of markets, that was roughly true. Word of mouth, local presence and time in the trade generally translated into being the obvious name customers thought of first.

That link has weakened. Discovery has moved to platforms with their own selection logic. Reputation earned offline doesn’t automatically transfer onto a comparison site’s ranking, a marketplace’s shortlist, or whatever an AI system decides counts as a good answer. An operator can be genuinely excellent and structurally invisible at the same time. Increasingly, that combination is common rather than rare.

This isn’t the same claim as “the internet matters now and it didn’t before.” Most established operators already have a website, a Google Business Profile, and a page of reviews under their own name. Those channels work exactly as well as they ever did. A search for the company’s name on Google still surfaces it, rating and history intact. That’s brand search, and it was never broken. What’s changed sits one step earlier, in category search: the query that never contains the company’s name, because the customer doesn’t have one yet. When a comparison site, a marketplace, or an AI assistant answers “best international movers from X to Y,” it isn’t retrieving a name someone already typed. It’s assembling one from whatever structured signals it can find. A reputation that lives in a Google Business Profile nobody searched for by name doesn’t get pulled into that assembly just because it exists.

Why this should worry established operators specifically

A reputation nobody can find is a reputation that isn’t working.

With nothing but a well-optimized listing and a handful of reviews, newer, smaller competitors are often better positioned inside these discovery layers than an established firm that’s never had to think about it. They built for the current path to the customer. The older firm built for the one that used to work, and being good at the work was never going to close that gap on its own.

Two operators are competing for the same shortlist slot on the same comparison platform. The first has been moving households and businesses for four decades: a real claims record and a loyal referral base. Forty years of goodwill like that never got entered into a system a machine could read. Its profile on the comparison platform has sat untouched since a junior staffer set it up three years ago: eleven reviews, the most recent one eight months old, a service-area field that still lists one country instead of the six the company actually covers. The second operator has been trading for eighteen months. It has no back catalogue of reputation to speak of, but its founder spent a weekend filling in every field the platform offers: a complete service-area map, licensing details entered where the platform has a field for them, and a review-request text that goes out to every customer within 48 hours of delivery. That process has produced forty-one reviews in eighteen months, all but three of them from the last four months.

On the platform’s own ranking logic, the newer operator wins that shortlist slot most of the time. Not because its reviews say anything more flattering (both companies average close to 4.7 out of 5), but because the platform’s ranking logic weights profile completeness and how active the listing looks, and the older company’s account has neither. No major moving-specific comparison platform publishes the exact weighting behind its ranking. This piece doesn’t pretend otherwise. What is publicly documented comes from comparable marketplaces that are more forthcoming about their own logic: Yelp tells businesses directly that an accurate, complete profile and a steady base of reviews are inputs to how a listing ranks, not just background detail. A moving-specific platform has no obvious reason to run on a fundamentally different logic. A platform built to surface current, well-documented listings has no mechanism for crediting forty years of work that was never entered anywhere it can see.

That’s the uncomfortable version of the problem. The reassuring version is that it’s fixable, and the operators with real reputations to bring already have the harder half done. What’s missing is rarely the substance. It’s making that substance visible inside the systems customers now use to choose.

How the discovery layer actually decides

Being “structured correctly” is the phrase doing the real work here, and it means nothing until it’s tied to fields a machine can actually read.

A comparison platform, a marketplace, and an AI assistant all run some version of the same underlying process: matching a query to a shortlist using whatever structured data is available, then ranking that shortlist by whatever signals predict a good outcome for their own business, usually a booked lead or a satisfied user, not a fair accounting of forty years of service. The signals that matter tend to be mundane rather than mysterious. Name, address, and phone number must match exactly across every listing and every mention of the company online, because a mismatch reads to most of these systems as an unverified or possibly defunct business. The service-area field needs to be complete, not a placeholder. Licensing and accreditation data belong wherever the platform has a field for them, since a field left blank behaves identically to a company with no accreditation at all, whether or not that’s true. Two of the platforms behind these systems are explicit about it. Yelp’s own guidance to businesses states plainly that keeping listing information accurate and building a genuine base of reviews are direct inputs to ranking. Google says the same about its own Business Profile: more reviews and positive ratings help a business’s local ranking. Review recency matters on top of that, even where a platform won’t publish the exact weight it carries. To a system built to flag active, current listings, a five-star average that hasn’t moved in three years reads closer to a dormant business than a well-regarded one.

AI assistants raise a separate issue. When a customer asks an AI system for moving-company recommendations, the system isn’t calling the company or reading its brochure. It’s drawing on whatever indexed, structured content already exists about that company: schema markup on the company’s own site, aggregator and directory listings, review-platform data, and increasingly the same comparison-site feeds a human shopper would see. Google publishes the exact specification its systems read for a local business, which makes this a documented requirement rather than a guess about how the tools behave. When a company’s website carries no schema.org markup identifying it as a moving business, no structured service-area data, and no machine-parseable accreditation information, it isn’t being unfairly excluded from that answer. It’s functionally invisible to the tool answering the query, in the same literal sense that a business with no listed phone number is invisible to someone trying to call it.

Why paying for placement doesn’t fix it

The obvious response, for an operator with the budget, is to buy the visibility instead of earning it structurally: pay for the featured slot, the sponsored listing, the top-of-page placement most of these platforms sell. It’s a reasonable instinct, and it isn’t wrong. Paid placement does put a listing in front of more customers than an unoptimized free one would reach on its own.

It doesn’t fix the underlying problem, for two reasons that both matter. First, paid placement resets the competition to a bidding war. An established operator doesn’t automatically win that war just because it has more revenue than an eighteen-month-old rival. A well-funded newer competitor, especially one backed by a marketplace’s own growth budget or a franchise rollout, can outbid an independent firm that’s never had to compete on cost-per-click before. Second, and less obvious: most of these platforms don’t treat a paid placement as a flat guarantee of visibility. They treat it as a boosted starting position inside a ranking that still weights conversion. The clearest documentation of that logic sits where the auction is biggest: Google states plainly that ad position is set by bid and auction-time quality together, not by bid alone. A marketplace selling its own sponsored slots has the same reason to protect the same thing. A listing can convert poorly for several reasons: a thin profile, a slow response time, or bad service-area data. That listing gets throttled even after the placement fee has cleared, because the platform’s own incentive is to keep sending customers toward listings that convert, not toward the ones that merely paid. An operator can spend real money on visibility and still lose the shortlist slot within a few weeks if nothing about the profile underneath that spend gives the platform a reason to keep showing it.

Paid placement can buy a temporary fix, or a fast start while the structural work gets done. It isn’t a substitute for that work. An operator that treats it as one ends up running a permanent, unwinnable bidding war against competitors with no back catalogue to protect and no reason to stop bidding.

Check what the machines can see

An operator can’t tell from the inside which of those two companies its own business currently resembles. The evidence sits outside the firm, in records it doesn’t control and has mostly never opened. Three checks settle the question, and none of them needs an agency or a budget.

The first covers the company’s own website. Google publishes the tool that reports what its systems extract from a page: a URL entered into the Rich Results Test returns whatever structured data that page exposes, which for a site that has never had any added is nothing at all. The specification behind that result rewards a slower read. Name and address are the only required properties for a local business. Telephone, opening hours, coordinates and aggregate rating are filed as recommended rather than required. A service area appears in neither list. The vocabulary exists, since schema.org defines an areaServed property, but a mover can mark up the six countries it covers in perfectly valid markup and still find that the fact its customers search on carries no documented weight in the rich result Google chooses to build. That gap rarely travels alone: pricing a system can’t read and a phone number that disagrees with itself usually sit on the same site.

The second check looks for duplicate profiles, and this is where long-established firms are most exposed. Google’s guidance to businesses is blunt: there should be one profile per business, and more than one causes problems with how the information displays across Search and Maps. Operators collect duplicates without noticing. A depot relocates, a franchise partner opens a second profile, an acquired company’s old listing never gets merged, and two records carrying different addresses and different phone numbers end up competing to represent the same firm. Searching the company name against each city it serves, one city at a time, usually surfaces them.

The third takes ten minutes. On every platform the company appears on, the date of the most recent review matters more than the total count. A profile carrying three hundred reviews and nothing since 2023 reads as the older operator from the scenario above, whatever the average score says.

Run together, the three produce a list of specific broken records instead of a general sense that the company ought to be doing more online. That difference decides whether the work ever gets done. Forty years of goodwill can’t be entered into any of these fields. Everything else on the list can.

Getting into the shortlist

The fix is being structured correctly for how modern discovery works: accurate, complete, machine-readable information about who the business is and what it does, not just a reputation that lives in people’s memories and old reviews. The operators who solve that stop being invisible to the exact systems now standing between them and the customers they’d win on merit alone.

None of this requires the company to look, sound, or operate differently. It requires someone to work through the list those three checks produce, platform by platform, including the several the company may not remember signing up for. The review process is the piece most often skipped, and it matters more than it looks. Asking every customer for a review in the days right after delivery, rather than hoping satisfied customers volunteer one unprompted, produces the steady trickle of recent entries that reads as an active listing. A large but static pile from three years ago doesn’t. It’s the same completeness-and-activity logic these platforms already apply to everything else on a profile. The fastest lever available to a long-established operator isn’t accumulating more reputation. The company already holds more of that than it needs. It’s making the reputation it already has visible on a rolling, current basis instead of leaving a one-time listing nobody has touched since it was created.

When operators treat this as a communications problem, a matter of writing better copy or taking better photos, they tend to see the least change from the effort. When operators treat it as a data-accuracy and maintenance problem, the same category of work as keeping a fleet’s insurance current, they tend to see the shortlist behavior shift, because that’s the category of signal these systems are reading.

Fixing the visibility side is real work, but it’s bounded work: a data-accuracy project with a clear end state, not an open-ended rebuild of the business.

Reputation is only half the picture. The other half plays out once a customer finds the company, and depends on whether they end up dealing with that company’s brand or someone else’s: see who actually keeps the credit for a job well done.

Every age invents its own proof that a stranger can be trusted before you’ve met them. A guild mark stamped into hammered metal. A wax seal on a merchant’s letter of introduction, carried three hundred miles ahead of the trader himself. A schema tag sitting in a webpage’s source code, invisible to anyone but the machine reading it on a customer’s behalf. The format keeps changing. What decides who gets trusted at a distance is whether a business bothered to speak the proof format its own era actually reads.

A tenant farmer could work the same land his entire life and still own nothing when he died: not the harvest, not the soil, not even credit for the labor, which the landlord’s ledger recorded under someone else’s name. Company towns ran an updated version of the identical trick a few generations later, paying wages in scrip redeemable only at the company store, so a worker’s whole earned livelihood stayed legible only inside a system he never controlled. Whoever keeps the ledger keeps the leverage, no matter what currency the ledger is written in.

Relocately’s homepage carries a number it clearly wants you to see: 5,000-plus reviews, from 175,000 users, across a network of 600-plus certified partners. It’s a good number. It’s also not your number, even on the job your crew did, on the truck your business owns, with the reputation your business built over years.

Nobody in the transaction ever forces the question of who earned it. The lead arrives, the job gets booked, the customer is happy, the invoice clears. The platform already knows the answer, which is exactly why it puts the number on the homepage and not on yours.

Who actually earned that review

The customer who leaves a five-star review after a smooth move is rating the experience they had. Someone showed up on time, handled their belongings carefully, communicated well, delivered what was promised. That’s real work, done by a real operator’s real people.

But the review doesn’t attach to the operator who did the job. It attaches to the platform that sold the lead. Every satisfied customer a marketplace routes to an operator becomes another data point in the marketplace’s own trust story, not the mover’s. Doing the job well enough and often enough builds someone else’s brand with the operator’s own labour.

The arithmetic on an ordinary month shows how fast that adds up. Say a mid-sized crew books 15 jobs a month through a single marketplace, and roughly a third of satisfied customers leave a review without being chased for one, which is a normal response rate for a well-run move. That’s 5 reviews a month, 60 a year. After three years of steady, well-handled jobs, the operator has quietly handed over close to 180 five-star reviews, searchable under the platform’s business name and feeding the platform’s own 5,000-plus number. The operator’s own Google Business Profile never saw a single one of those reviews. It’s the profile an actual future customer might search for by name.

The reviews you don’t own

Some marketplaces do show a version of your name. A profile page, a rating specific to your business, a handful of quotes pulled from real reviews. It looks like credit. It functions differently.

That profile lives on a URL the marketplace controls, ranked by an algorithm the marketplace controls, on a domain that captures the search traffic for your business name before your own site gets the chance. When you stop paying for placement, or leave the platform for a competitor, the profile with your name on it typically stops surfacing too: not deleted, necessarily, just buried under whichever operator is paying for that slot now. Moving marketplaces don’t publish their own ranking mechanics, but the pattern is well documented one tier over, in the home-services lead marketplace that pioneered this exact model. Angi’s own FAQ states plainly that a pro’s choice to advertise “does not affect their ratings or reviews,” confirming the review data survives while a separate, paid mechanism decides whether a customer ever sees the profile it sits on. The reviews themselves might still sit in the platform’s database somewhere. They stop doing any work for you the moment you’re not the one paying to be found.

A review sitting on your own Google Business Profile behaves differently. It’s discoverable by your business name specifically, not surfaced only inside a category search that fifty other operators are also bidding into. It’s still there in five years whether or not you bought a single lead that month. It compounds toward something you own: higher local search visibility, more direct-search bookings, a lower cost to win the next customer. A platform-hosted review compounds too. Just not for you.

None of this is an argument for walking away from marketplace demand, which for a lot of operators is a legitimate, even necessary, part of a healthy pipeline. It’s an argument for noticing that lead volume and reputation equity are two different assets, and only one of them is being paid into an account you control. A marketplace lead can hand you this month’s job. It very rarely hands you next year’s cheaper one.

Reputation only counts once you’re in the room

An operator with decades of history, genuine accreditations, a strong claims process and real destination-agent relationships has something worth a great deal, but only once a customer knows that operator exists. Increasingly, the path to that customer runs through an intermediary first: a comparison site, a marketplace, a directory. Sirelo alone lists more than 26,000 movers and says it connected over 200,000 consumers with removal companies in 2025.

An operator discovered through one of those competes on the terms the platform sets: usually price, usually against several other quotes, usually without the customer ever learning much about who they are beyond a name in a list. Forty years of doing the job right doesn’t show up in that comparison. It can’t. The comparison was never designed to carry it.

The comparison table a customer sees makes this concrete: four movers quote the same three-bedroom relocation at $4,200, $4,450, $4,600, and $4,900. The $4,900 quote belongs to the operator with the longest track record, the best claims ratio, and the only crew on the list that’s handled that specific corridor a hundred times before. None of that is a column in the table. Price is the column. Absent any other signal, the customer’s eye goes to $4,200 first. The platform’s own copy nudges them there too, since a cheaper headline quote keeps its own conversion numbers up. The strongest operator in the list is competing on the one dimension where being the strongest operator counts for the least.

What the aggregate hides

The marketplace’s big number buries a trade-off: it cuts against the operator who’s good at the job.

A platform’s aggregate score, the 5,000-plus reviews, the 175,000 users, smooths over every individual operator’s variance. That’s genuinely useful to a customer who has no other way to screen forty movers in an afternoon; it’s why the aggregate exists and why it converts leads into bookings. But it means a stellar operator’s track record and a mediocre one’s both get folded into the same headline number a customer sees first. The strong operator doesn’t get a materially better shot at next week’s lead because of last month’s five perfect reviews. The weak operator doesn’t get meaningfully penalized either, as long as the platform’s overall average holds up. Under the same marketplace banner, from a customer’s first glance, the two are roughly interchangeable.

That’s not a flaw in how marketplaces work. The mechanism works exactly as built: it reduces a customer’s decision to a manageable comparison in the time it takes to fill out one form. The trade-off is who benefits from that simplification and who pays for it. A weaker operator gets a floor they haven’t earned. A stronger one gets a ceiling they have.

Why the platform needs you replaceable

This isn’t an accident of design. The model works the way it has to work to be worth building.

A marketplace’s power over the operators supplying it depends on those operators staying replaceable. If reviews, relationships, and referral demand accumulated visibly against an individual operator’s own name inside the platform, the best operators would eventually build enough standing with customers to quote direct next time and skip the fee entirely. The platform’s value to the customer and its pricing power over the operator rest on the same structural fact: the customer trusts the marketplace’s aggregate more than any single mover’s specific record. When that trust stays pointed at the platform level, every operator underneath it stays a fungible input: easy to compare, easy to swap out, easy to charge the same fee whether the job was excellent or merely adequate.

This isn’t unique to moving, and it isn’t a conspiracy. The same logic shows up anywhere a platform sits between a service provider and the end customer: a driver’s rating usually stays with the app that collected it rather than following the driver anywhere else. A restaurant’s reviews usually live on the delivery app’s own listing before they live on the restaurant’s own site. Whichever party owns the customer relationship at the moment of the transaction has a structural reason to keep owning it afterward too, not because it’s dishonest, but because it’s the whole basis of the platform’s pricing power. An operator who understands this stops being surprised the review stayed behind. It was never going to travel. Its job was to keep working for whoever collected it.

The quiet trade every operator is making

Every job fulfilled through someone else’s platform pays that platform’s reputation and your bills, in that order.

Marketplace work itself isn’t the problem. For plenty of operators, it’s a legitimate part of a healthy pipeline. The problem starts when it becomes the whole pipeline, when every job an operator books runs through a platform’s storefront. It builds nothing the operator will still own next year.

A review, a repeat customer, a referral: these compound. They make the next sale easier and cheaper than the last one. That compounding stops entirely if the customer never meets the operator’s own brand. It happens for the platform instead, deal after deal, at the operator’s expense.

Two operators on the same starting line show what that gap costs over time. Both run 150 jobs a year at a $2,800 average ticket. Both start with no meaningful brand recognition in their market. Operator A routes nearly all of it through marketplace leads, paying a per-lead or per-booking fee on every single job, indefinitely, because nothing about the arrangement lowers the acquisition cost of job 500 against job 1. Operator B routes the same volume through marketplaces at first, but treats every completed job as a chance to capture the relationship: a review request that lands on the operator’s own Google Business Profile, a follow-up that comes from the operator’s name and number, a referral ask at the moment satisfaction is highest. By year three, even a modest 20% of Operator B’s volume comes from repeat customers and referrals at close to zero incremental acquisition cost. Operator A is still paying full freight on job 500, exactly like job 1. Same revenue line on both spreadsheets. A very different business sitting underneath it.

Keeping the brand in front

The fix isn’t refusing marketplace demand. It’s making sure the customer deals with, remembers, and reviews your own brand. The infrastructure underneath can run whether or not you built it yourself: the qualification, the follow-up, the systems that used to require an internal digital team.

Concretely, that means the confirmation message after booking comes from your business, not the platform’s. It means the review request goes out under your name, timed to the moment the crew finishes unloading, pointed at your own Google Business Profile instead of a feedback form that only feeds the platform’s dashboard. It means the follow-up call three weeks later, checking that nothing arrived damaged, comes from a number the customer already recognizes as yours, not a marketplace support line. None of that requires the job to have started as a direct search. It requires only that the moment the job is won, the relationship stops belonging to whoever sold the lead and starts belonging to whoever actually did the work.

The honest objection is time. Building a direct channel from nothing while marketplace leads are already flowing in feels like extra work with no schedule of its own. It doesn’t have to start that way. The relationship-capture layer above attaches to jobs you’re already booking through a marketplace today: confirmation under your name, review request under your name, follow-up under your name. It doesn’t require replacing marketplace volume before it starts paying off. It requires only that the next satisfied customer, wherever the lead came from, leaves knowing your name and not just the platform’s.

The name on the invoice

This is one half of a bigger problem. The other half happens before a customer ever sees an operator’s reviews at all: read what a strong reputation is actually worth if nobody finds you.

Between the two, the pattern repeats. Decades of good work only count for the business whose name is on the invoice the customer keeps. Right now, for most operators taking marketplace volume, that name isn’t theirs. Reviews compound. Referrals compound. A rating a customer can find under an operator’s own business name compounds. Capturing any of it isn’t complicated. It has to be built somewhere other than someone else’s storefront, one job at a time, starting with the next one that comes through the door.