What a Landing Page Should Ask For, and What It Should Never Ask First
A form is triage, not an interview. Ask only what you need in order to reply usefully, give something before you ask for anything, and add a field only when it changes the reply.
On This Page
The form asks for eleven things. The visitor arrived with one question. That mismatch is the entire problem, and it is not solved by moving the button, changing its colour, or writing a livelier label above the first field.
A form is triage. Its job is to collect exactly what you need in order to give a useful reply, and nothing else. Not what the sales team would enjoy knowing. Not what the customer database has a column for. Not what a competitor asks. What you need in order to answer well, and fast enough that the answer still matters.
Almost every landing page conversion argument resolves itself once that is settled: how many fields, in what order, worded how, with what above them and what after the button. Teams argue about those things for weeks because they never agreed on what the form is for.
This piece covers the sequencing of asking and giving, the real price of each field, the difference between qualifying somebody and interrogating them, and the one condition under which a long form is the correct answer.
Triage, in the ordinary sense
Triage is not a full history. It is the smallest set of questions that decides what happens next and how urgently. Everything else gets asked later, by the person who is now actually dealing with you. A form that tries to be the whole intake process instead of the first thirty seconds of it is doing the wrong job.
There is a simple test for each field. Finish this sentence: without this answer, my reply would be worse because… If you cannot finish it without inventing a reason, delete the field. If you can finish it, keep it, and make sure the visitor could work out why you asked. Our parent company set out the same idea in a form is a triage instrument, and it holds for enquiry forms, quote requests and support intake alike.
The test is uncomfortable because most forms fail it in four or five places at once. That is the point. A form is one of the very few places where a marketing team asks strangers to do unpaid work, and the work should be proportionate to what they get back.
What a field actually costs
Fields get added one at a time, each with a reasonable-sounding argument, and nobody prices them. Every field costs three separate things, and only the first one is ever discussed.
Three costs, not one
The obvious cost is attention: one more thing to read, decide about and type on a phone, on a train, with one hand. The second is hesitation. A question the visitor cannot answer confidently, such as a budget band before they know what the work involves, does not merely slow them down. It gives them a reason to leave and think about it, and that is a decision they rarely come back from.
The third cost arrives after submission. Every field is a record you now hold, store, back up, share with whichever tools are connected, and eventually have to explain if somebody asks what you keep about them. Data minimisation is not only a legal principle. It is a way of not accumulating obligations you had no use for.
| Field | What it buys you | When to keep it |
|---|---|---|
| Name | A reply that reads as though a person wrote it | Always, as one field, never split into three |
| Email address | The ability to reply at all | Always |
| What they need, in their words | The only input that shapes your answer | Always, and give it real room |
| Company or organisation | Context your first reply would otherwise spend a paragraph asking for | When your answer genuinely changes by organisation type |
| Job title | A guess at who decides | When the reply differs by role, which is rarer than assumed |
| Budget range | A sorting rule for your pipeline | Rarely, and almost never before a conversation |
| Phone number | A faster channel for you, not for them | Only when the visitor has asked to be called |
| How did you hear about us | A weak attribution signal | Only when nothing else in your measurement can answer it |
For every field, imagine the visitor asking why you want it, and answer out loud. "So we can send this to the right person" is a good answer. "So the pipeline report has a column" is not, and it is the real reason more often than anyone admits. Fields that fail this test are also the ones most likely to be filled in inaccurately, which makes the data worse as well as the experience.
Give first, then ask
The order on the page matters as much as the fields in it. A visitor who still does not know what you do, whether you have done it before, or roughly how you work is being asked to pay before seeing the menu. They will not, and the form will be blamed for a problem that sits above it.
So sequence it: state the claim, show what backs it, be specific about the thing itself, then ask. And when a page has no worthwhile offer at the end of that sequence, the arrangement of the form is irrelevant. That is the argument in the offer is the campaign, and it is worth reading before anybody spends another fortnight testing button copy on a page whose proposition is the actual failure.
Giving is more literal than it sounds. Name the thing you actually do, in the words a buyer would use. Show one piece of work at the level of detail a sceptic would want. Say what happens after somebody gets in touch, including who they will hear from. None of that is persuasion in the theatrical sense. It is the information a reasonable person needs before handing over their name, and pages that skip it are asking on credit.
It also helps to stop treating the form as the only door. Some visitors have a question the site could answer immediately, and forcing them through a contact form to get it is a poor trade for both sides. A widget like Flidu takes that seriously: it answers from the site’s own content, and offers contact and conversion actions to the people who still want a human. The form then receives fewer enquiries and better ones, which is the outcome most teams were trying to buy with extra fields.
Qualification or interrogation
These two get confused constantly, and the difference is easy to state. A qualifying question changes what you do next. An interrogating question changes only how you file the person afterwards.
Questions that genuinely qualify
A question qualifies when two different answers send the enquiry down two different paths. Which product line this concerns, whether an existing site is involved, whether there is a fixed date to work back from: each of those redirects who replies and what the reply contains. One or two of them on a form is reasonable, because the visitor gets a better answer in exchange.
Questions that only sort
Budget bands, headcount dropdowns and timeline pickers usually sort rather than qualify. They also invite fiction: people pick the middle option, or the one they think will get them a reply. The honest place for those questions is the first conversation, where you can explain why you are asking and what each answer would mean. Asking them cold, of a stranger, mostly produces confident-looking data that nobody should trust.
There is a predictable moment when these fields multiply. Spend rises, the mix of enquiries gets worse, and somebody proposes adding qualifying questions to hold the line. It is the wrong lever, for the reasons set out in why paid search stops working when you scale the budget: the enquiries got worse upstream, in targeting and in the offer, and a longer form only hides the symptom while quietly taxing the good enquiries too. Fix it in the campaign, not in the form.
If the answer to a question would not change your reply, you are not qualifying anybody. You are charging them a fee to contact you.
The minimum viable ask
For most enquiry forms on most sites, this is the whole thing, in this order.
- Who you are. One name field. Splitting it into first and last is a database preference imposed on a person, and it buys you nothing you cannot derive later.
- Where to reply. An email address, with a sensible input type so phone keyboards behave, and validation that explains the problem instead of colouring the border red.
- What they need, in their own words. A generous open field, not a 200-character box. This is the only answer that reliably improves your reply, so give it the space that implies.
- One thing that changes the route, if anything does. A single conditional field at most. If you need two, one of them is probably serving a report rather than the person.
- Plain consent, only if you will do more than reply. Replying to an enquiry needs no marketing permission. Adding somebody to a mailing list does, and it should be a deliberate opt-in rather than a pre-ticked afterthought.
Everything else is a design and accessibility question: visible labels that stay visible, errors announced rather than merely shown, touch targets people can actually hit, and a keyboard path that works. Those are not extras. They are the difference between a form that looks finished and one that is, and they belong in interface and product design rather than on a post-launch fix list.
When a longer form is the right answer
Now the concession, because the short-form position gets argued well past the point where it is true. Long forms are correct in a specific set of cases, and they share one property: the visitor already understands that answering is the process, not a toll on the way to it.
- Selection is the product. Applications, grant submissions, waiting lists with limited places. Answering is the work, and a short form would misrepresent what happens next.
- A quote is impossible without inputs. Dimensions, volumes, locations, regulated details. Ask for them, and show why each one changes the number.
- The request has real constraints. Dates, capacity, delivery areas. A booking that ignores its constraints creates a worse experience than the questions would have.
- Support intake, where a vague ticket helps nobody. Version, environment, order reference. The person asking wants the fix, and they can see that these answers speed it up.
- Your bottleneck is capacity, not demand. If the team cannot answer everything well, fewer and better-matched enquiries is a rational trade. Make that a decision somebody signs off, not an accident of form design.
That last case is where most of the honest disagreement lives. A longer form does tend to raise the average quality of what arrives, and if your constraint is people rather than pipeline, that can be the right choice. Just measure it properly, because the count of submissions will fall and somebody will read that as failure. Cost per lead is a trap covers why that metric misleads here more than almost anywhere else: cheaper enquiries that never convert look like a win for months.
The part after the button
The form is not the product. The reply is. A visitor who sends a careful description of their situation and receives an automated acknowledgement with no human follow-up for four days has learned something accurate about how you work, and it is not what the page promised.
So the confirmation screen should say three things: what you received, who will respond, and roughly when. Vague reassurance is worse than nothing, because it sets an expectation you did not choose and cannot meet.
If the message failed to send, say so on the screen and offer another route immediately. A green tick over a lost enquiry is the most expensive bug a marketing site can carry, because nobody reports it: the visitor believes they contacted you, and you never knew they tried.
Behind the screen, the submission should route itself to the right person, carry its context with it, and start a record you can still read in a year. That is ordinary marketing automation work, and it is also where measurement begins: the fields you captured, the source you recorded and the outcome you eventually wrote back are what let you build attribution you can defend rather than a story assembled after the fact. A form designed only for the visitor and never for the record produces a pleasant experience and an unanswerable question six months later.
What we would change first
Take your current form and cut it to four fields for one month. Then measure the thing that matters, which is not submissions: count the enquiries that turned into a real conversation, and read what people wrote in the open field. That reading is the fastest conversion work available to most teams, and it costs nothing but attention.
If the shorter form brings in more enquiries than you can serve, you have found a real constraint, and you can add fields back deliberately, one at a time, each with a stated reason. That is a different situation from starting with eleven fields because nobody was sure which ones mattered.
And if you are unsure which of your fields survive the test at the top of this piece, print the form, put a line through every field you cannot justify out loud, and see what is left. Usually it is four things and a sentence. If you want a second opinion on yours, send us the page and tell us what you do with each answer.
Common questions.
How many fields should a landing page form have?
As few as let you reply usefully, which for most enquiry forms means a name, an email address and an open description of what the person needs. Add a field only when two different answers would send the enquiry down two different paths. Fields that exist to populate a report, rather than to shape your response, cost you enquiries and produce unreliable data at the same time.
Should a landing page form ask for a phone number?
Only when the visitor has chosen to be called. A phone field is a faster channel for the business and a slower, more exposed one for the person filling it in, so making it mandatory reliably reduces completions. If calling is genuinely part of your service, offer it as an option the visitor selects, and make the field appear once they have asked for that route.
Do budget questions improve lead quality?
Rarely, because they sort rather than qualify. A stranger who does not yet know what the work involves cannot answer a budget question accurately, so they guess, pick the middle option, or leave. The same question asked in a first conversation, with context and an explanation of what each range would mean, produces an answer you can actually plan against.
Is a multi-step form better than one long form?
Splitting a form into steps helps when the questions genuinely belong in stages and each step shows progress, such as a booking that narrows by date and then by option. It does not help when it simply hides the same twelve fields behind three screens. If the total ask is unjustified, steps make it feel longer rather than shorter, and abandonment moves rather than falling.
What should a confirmation page say after someone submits a form?
It should state what you received, who will respond and roughly when, in plain language. Avoid vague reassurance, which sets an expectation you did not choose. If delivery failed, the screen must say so and offer another way to reach you immediately, because a success message over a lost enquiry is invisible to you and permanent for the person who sent it.
When is a long application form the right choice?
When answering is the process rather than an obstacle before it. Applications, grants, waiting lists with limited places, regulated quotes and support intake all justify detailed questions, because the person can see that each answer changes the outcome they came for. The test is whether the visitor would understand why you are asking, without you having to explain it afterwards.
Facing this in your
own business?
Tell us where you’re headed — we’ll map the shortest honest route.