We recently sat down and reviewed a stack of real, live moving-company websites, quote forms included. These weren’t neglected pages thrown up a decade ago and forgotten. Several belonged to established operators who clearly invest in their web presence: current design, professional photography, years of blog posts. That’s the detail worth sitting with. The businesses running the best-looking sites in the group were committing the identical mistake as the businesses running the worst ones.
“Why does everyone start my quote form, and almost nobody finish it?”
That’s the question behind this piece, and it isn’t really a design question. A flight-search page and a moving-company quote form are structurally the same kind of instrument: a form standing between a stranger and a price. The gap isn’t that airline search sites hired better designers. A flight search asks for one thing, in the order the traveler already expects, and nothing else. A moving-company quote form almost never does.
Why does everyone start and almost nobody finish?
Three defects showed up on form after form, independent of how new the site was or how much the operator had obviously spent on it: fields front-loaded before the visitor has any reason to trust the business, questions that serve the operator’s records rather than the visitor’s task, and interruptions dropped into the middle of a form the visitor was already filling out. None are exotic. Each is a known, well-documented usability problem. It sits in plain sight on pages built by people who would tell you, correctly, that they take their website seriously.
The mechanism isn’t complicated. Why does a moving-company quote form ask for square footage, move date, origin, destination, bedroom count, inventory type, preferred contact time, and “how did you hear about us” on one screen? The honest answer is rarely “because a visitor wants to hand over all of that before seeing a number.” The form was built to satisfy whatever sits on the other end of it: sales wants a complete record before the first call, marketing wants an attribution field, and nobody built it to represent the person filling it out. A flight search doesn’t have that problem, because nothing downstream of the search box is asking it to collect a passport number before it shows a price. It asks for the one decision the traveler is ready to make, in the order they’re already thinking in, and defers everything else until it’s needed.
The obvious objection belongs here: “We need all this information to give an accurate quote, so we ask for it all up front.” It sounds reasonable, and it doesn’t survive contact with what an accurate quote actually requires. A flight search doesn’t need a seat preference to show a price; it needs origin, destination, and dates. A moving quote doesn’t need to know how the visitor heard about the company to estimate a job; it needs the same handful of data points, in roughly the same order. Everything past that is a record the business wants, not a number the visitor is waiting on.
The three specific defects
Front-loaded fields. Luke Wroblewski cites a comparison of two Contact Us forms: when an 11-field version was replaced with a 4-field version, submissions rose 160% and conversion rose 120%, and the quality of submissions stayed the same. The same piece reports a major online travel site that removed an optional “Company” field customers mistook for a request for their bank details, and saw an overnight increase of $12 million a year in profit. Baymard Institute’s benchmarking found that 39% of the sites in its e-commerce benchmark require a phone number with no explanation for why, and its survey of 1,026 online shoppers found that 14% would never give an online store their phone number and 27% are reluctant to give a date of birth. Not one of the moving-company forms reviewed explained a required field. Every one of them just asked.
Dead-weight questions. “How did you hear about us” is the clearest example: attribution data the marketing team wants, asked at the exact moment the visitor is deciding whether continuing is worth their time. A form field is a sentence the visitor has to read and answer before moving forward. If a field doesn’t change the number the visitor is about to see, it doesn’t belong on the screen where they’re deciding whether to keep going. Attribution can be captured later, anywhere that isn’t the moment the business needs the visitor’s patience most.
Mid-flow interruption. One site reviewed does something more aggressive than either problem above: it interrupts an in-progress quote form with an autoplaying video of the company’s history, unrelated to the fields the visitor was filling in. Nielsen Norman Group’s research on disruptive workflow design is built around a different scenario: an app-update interstitial that forces a user to restart a task rather than resume it. The principle carries over: an interruption dropped into an active task can read to a visitor as a reason to leave, not something to tolerate. A flight search has never once asked a traveler to sit through a video about the airline’s founding between entering a departure city and a destination. Anything unrelated, inserted mid-task, asks the visitor to do the business’s marketing a favor at the exact point they’re least willing to. The video itself isn’t the problem. It just needs a place on the page that’s actually earned, not a spot in the middle of someone else’s task.
These aren’t the only ways a form loses someone. A form can ask the right questions in the right order and still fail if a search engine or a CRM integration can’t read it: that’s the quote that dies in the form, a separate problem beneath this one. This piece is about what happens once a human, not a machine, is staring at the fields.
What staged, relevant disclosure actually looks like
The fix isn’t adding steps. The easy, wrong takeaway from “don’t front-load everything” is “build a multi-step wizard instead,” and that isn’t what the evidence supports.
Jakob Nielsen described the underlying pattern, progressive disclosure, in 2006: defer anything advanced or not yet needed to a later screen, so the first thing a person sees is only what they need to make the next decision. NN/g’s later research on workflow timing sharpens the stakes. In its testing of B2B sites, business professionals usually rebelled when a site collected lead information too early, and the result was “no leads at all, as users proceed to more welcoming sites.” A flight search is a working example of both ideas at once. Screen one asks where and when; nothing about seat class or payment details appears until the traveler has already seen prices and picked a flight worth paying for.
A moving-company quote form built the same way would ask, first, only what a rough estimate needs: origin, destination, approximate size, and a move date, the same four a flight search opens with. Once the visitor has seen a price range, and has a reason to believe the business is worth five more minutes, the form can ask what refines that number: inventory specifics, access details, additional stops, preferred contact time. Contact information comes last, even though it’s the field every operator is most tempted to ask for first. By that point, the visitor has already invested enough attention that finishing feels like the easier choice.
This next idea is easy to misapply. Whether the staging happens across three screens or inside one well-ordered page isn’t the variable that decides the outcome. Wroblewski’s conclusion is blunt: whether a form is single-page, multi-page, or dynamically staged, “in tests I’ve seen it doesn’t make a difference.” Baymard’s own research into one-page checkouts found the same thing differently: the A/B tests showing a single-page form winning often compare a badly built multi-step flow against a newly optimized single page, not a fair fight between two well-built versions. The friction Baymard finds in usability testing comes from what a person has to do at each step, not how many steps exist. A 2021 study in JMIR Human Factors had 20 hospital staff test single-page, multi-page, and conversational versions of a patient-referral form and found the single-page version scoring highest on usability and finishing fastest. Format isn’t irrelevant everywhere. But it’s one small study in a context that isn’t a moving quote, and it doesn’t license the broader claim that steps are the fix. The defensible version of this idea is narrower than “add more screens”: ask for only what’s relevant at the moment you ask for it, in the order the visitor already expects, and explain anything unusual. That’s the whole rule.
Fixing the generic error message
A fourth defect shows up even on forms that get the field order right: the generic error message. “This field is required.” “Invalid entry.” Text in red, no explanation of what’s wrong.
Nielsen Norman Group’s ninth usability heuristic exists for this. Its guidance is unambiguous: “error messages should be expressed in plain language (no error codes), precisely indicate the problem, and constructively suggest a solution.” Google’s Material Design guidelines, from the product side, say to clearly explain the error and how to fix it, and to label error messages individually as the user works through the form. “This field is required” fails the first test. It doesn’t say what’s wrong with what was entered or how to fix it.
Underneath the UI framing, this is a writing problem. “This field is required” hasn’t said what is wrong. “Enter a valid US ZIP code, five digits” has. The rewrite isn’t longer or cleverer. It just names the actual problem and the actual fix, in words a stressed visitor can act on without a second read.
None of this is free, and it isn’t meant to read as one. Amazon didn’t remove checkout friction out of politeness: Amazon patented the single-click version of it in 1999. When that patent expired in 2017, industry coverage of the moment cited average cart abandonment sitting around 70%, the exact number every extra field and every vague error message quietly feeds. A visitor who gives up on your quote form may not vanish. Some resurface on a marketplace instead. That feeds the same mechanism behind what rented demand really costs. The form you already own is either winning that visitor directly, or handing them to a channel that charges you to win them back.
This fix has a ceiling, not just a floor. Rebuilding a quote form this way closes a real, current gap: most of the forms we reviewed hadn’t done it, and it doesn’t buy a permanent advantage. Once field order and plain-language errors become normal, as they may, that stops being what sets one operator apart. What stays defensible past that point isn’t the form’s mechanics; it’s what the operator does with the extra completions it now delivers. The corrected form is table stakes. The business built on top of it is where the real difference still lives.
So: why does everyone start your quote form, and almost nobody finish it? Your form abandonment rate isn’t a mystery once the cause is named, and it’s exactly the finish-rate half of the two numbers that actually tell you whether a website is working. Your form probably shares the defects we saw in the forms we reviewed: it asks for a complete record before earning the right to ask for anything, mixes the business’s own paperwork into the visitor’s task, and, on at least one site reviewed, interrupts the task altogether. None of it requires new software to fix. It requires asking, field by field, whether the person on the other end has any reason yet to answer. A flight search never asks a question it hasn’t earned. Neither should the form standing between your business and a customer’s first real look at your price.
