Look at the drop-off chart for almost any form and the shape is the same. A cliff at the start, then a gentle slope, then a small dip near the end. The middle behaves the way you would expect. The start does not.
The instinct is to read that cliff as a length problem. The form looks long, so people bail. That leads to the standard advice, which is to cut questions, and cutting questions does help a little. But it does not explain why a four-question form and a fourteen-question form often lose a similar share of people at the same point.
What is actually happening is that the first question is doing a job nobody designed it to do. It is answering a question the visitor has not asked out loud: what is this going to be like?
The first question is a sample
Nobody reads a form before starting it. They glance at the first thing and extrapolate. If question one is easy, the rest are probably easy. If question one is work, the rest are probably worse, and there are apparently eleven of them.
This is why the single most expensive thing you can put first is a long free-text box. It says, before anything else, that this form expects effort. The person has not typed a word and has already decided the answer is no.
It is also why asking for a phone number first is so costly. It is not a difficult question. It is a question about trust, asked before you have earned any, and people notice being asked for something valuable by a stranger.
People do not abandon forms because they are long. They abandon them because the first question implied the rest.
What a good opener looks like
The best opening question is one the person can answer instantly, without thinking, without typing, and without giving anything away. A single choice from a short list is close to ideal. Which service are you interested in. How many people. When do you need it by.
It works partly because it is easy and partly because it is relevant. It makes the form feel like it is about the visitor's problem rather than about your database. A form that opens with the service someone wants reads as a conversation. A form that opens with their full name reads as paperwork.
There is a commitment effect too, and it is real even if it gets overstated in marketing blogs. Somebody who has answered one question is measurably more likely to answer the second. The first answer costs you almost nothing and buys a disproportionate amount of momentum.
Move the identity questions later
Name, email, phone and company are the questions you actually want, which is exactly why they should not come first. They are the price of the form, and you charge a price after you have shown the value, not before.
Turn it around. Ask two or three easy, relevant questions, let the person see that the form understands what they need, and then ask for contact details at the point where giving them is obviously the way to get what they came for.
Most people who reach the last question of a form finish it. The hard part is getting them to the last question, and front-loading the identity fields is one of the most reliable ways to make sure they never do.
The label problem
Sometimes the opening question is fine and the wording around it is not. Field labels written for internal use leak into public forms constantly. Enquiry type. Lead source. Service category. Every one of those is a phrase from your CRM rather than a phrase anyone would say.
Read your first question out loud as though you were saying it to somebody across a counter. If it sounds like a form, rewrite it until it sounds like a person. What can we help with beats enquiry type, and costs nothing.
The same applies to the help text underneath. If it explains what the field means, the field is badly named. Fix the name and delete the explanation.
One question per screen changes the calculation
If your form shows one question at a time, the first question matters even more, because it is literally the entire form as far as the visitor can see. There is no context, no sense of length, nothing but that single question and a button.
This cuts both ways. A good first question in that format is extremely effective, because the form appears to be a single easy thing rather than a wall. A bad one is fatal, because there is nothing else on screen to soften it.
It also means you can be honest about length in a way a long page cannot. A progress indicator that says one of six is reassuring when the first of the six took two seconds. The same indicator on a form that opened with a paragraph box reads like a threat.
Required fields at the start are a trap
Marking the first question required is the default in most tools and it deserves more thought than it gets. A required first question means a person who is not sure how to answer it cannot proceed, cannot skip, and cannot see whether the rest of the form is any better.
If the question genuinely gates everything that follows, required is correct. If it is merely useful, let people move past it. You would rather have a submission missing one optional field than no submission at all, and in practice the vast majority of people answer optional questions anyway.
The exception is anything you cannot follow up without. One contact field, near the end, genuinely required. That is usually the entire list.
Watch a real person do it
Analytics tell you where people stop. They do not tell you why, and the why is usually visible within thirty seconds of watching somebody try.
Find one person who is not in your business, sit them in front of the form, and say nothing. The pause before they touch anything is the thing to watch for. If they hesitate, reread, or ask you what you mean by something, the first question is the problem and you have just found it without any instrumentation at all.
Do this with three people and you will have a better idea of what to change than a month of dashboards would give you. It is the cheapest research available to a small business and almost nobody does it.
A worked example
A removals company had an enquiry form that opened with a text field labelled moving from, then moving to, then a date, then name, email and phone. Completion sat a little under thirty per cent, and most of the loss was at the first field.
The change was small. The first question became a set of four buttons: studio, one bed, two bed, three or more. Nothing else moved. Addresses came second and third, contact details last.
Completion went to just over half. Nobody had removed a question, shortened the form, or made it prettier. They had simply put something at the front that a person could answer while still deciding whether to bother.
The checklist
Before you change anything else about a form that is losing people, run through this. It takes ten minutes and it resolves most of the cliff.
- Can the first question be answered in under two seconds, without typing?
- Is it about the visitor's problem rather than your records?
- Does it sound like something a person would say out loud?
- Are name, email and phone all below it rather than above it?
- If it is required, would you genuinely rather lose the submission than lose the answer?
- Would a stranger know what to do without reading any help text?
The button matters more than you think
Whatever sits next to your first question is part of the first question. A button that says Submit on a form with eight more questions behind it is a small lie, and people who click it and find themselves on question two have learned that the form is not being straight with them.
Next is honest and dull. Get my quote, See availability and Check my options are honest and useful, because they restate what the person is going to get in exchange for continuing. That restatement is doing real work at exactly the moment somebody is deciding whether to bother, which is why generic button text is such a cheap thing to fix.
Avoid anything that sounds like a commitment the person has not made yet. Start my application, on a form that is really an enquiry, will lose people who were perfectly happy to enquire and have no intention of applying to anything today.
What sits above the form is part of it
People do not experience the first question in isolation. They experience the heading above it, the paragraph next to it, and whatever the link they clicked led them to expect. If the page promises a quick quote and the first question asks for a project brief, the mismatch does the damage, not the question.
This is why forms sometimes perform very differently on two pages with identical fields. The traffic arriving is in a different state of mind, and the same opener reads as reasonable in one context and presumptuous in the other. Before you rewrite the question, read the sentence immediately above it and ask whether the two are describing the same thing.
A single line of expectation setting fixes a surprising share of first-question drop-off. Six quick questions, about a minute is a sentence that costs nothing and removes the fear that the form is endless, which is the fear most people are actually acting on.
Do not make people choose a category they do not know
A very common opener asks which of your services somebody wants, using your names for them. This is fine when the names are obvious and quietly disastrous when they are not. Somebody who cannot tell whether they need a compliance audit or a compliance review will not guess. They will leave, because guessing wrong feels like it will waste everybody's time.
If your categories require knowledge the visitor may not have, describe the situation rather than the service. What are you trying to sort out beats which service do you need, and it lets you do the categorising yourself once you have the answer, which you were going to do anyway.
The test is whether somebody who has never heard of your company could pick correctly on the first try. If the answer is no, the question is written for you rather than for them.
The opener you can reuse
If you want one shape that works across almost every kind of enquiry form, it is this: a short list of three to five options, describing situations rather than products, with a plainly worded final option for anything that does not fit.
It is answerable instantly, it is visibly about the person rather than about you, it gives you a routing signal you would otherwise have to ask for later, and it never traps anybody with no correct answer. There is a reason it turns up on almost every high-performing form once somebody has spent a year iterating on one.
Three to five is the range because two feels like a false choice and seven means reading. If you genuinely have nine services, group them into five and ask a follow-up, which costs one extra tap for the people who need it and nothing at all for everybody else.
What this does not fix
A good first question will not save a form that asks for twenty-five things nobody understands the purpose of. If your middle section is a slow bleed rather than a clean slope, that is a different problem and it needs the same scrutiny applied question by question.
It also will not fix a traffic problem. If the people arriving are not the people you want, a better form converts more of the wrong audience, which feels like progress for about a month.
But the cliff at the start is nearly always fixable, it is nearly always fixable this week, and it is nearly always worth more than anything else on the list.