Most B2B websites ask a visitor to fill in a form before anything else can happen. The demo request, the contact form, the gated document. Whatever sits behind it, the form is the point where interest either converts into a record in your CRM or disappears.
The Baymard Institute has spent more than a decade running usability testing on checkout and form flows. Their work is on e-commerce, not B2B, which is exactly why it is useful. E-commerce measures abandonment precisely, at scale, with money attached. The mechanics that make people give up on a form do not change because the form is asking for a demo instead of a delivery address.
In this article
What the research measures
Baymard’s checkout form field research tracks the number of fields real sites ask people to complete. The average in 2024 was 11.3 fields, down from 11.8 in 2021 and 12.7 in 2019.
Their conclusion from testing is that most sites need only 8 fields in total.
That gap, roughly three fields, is not a rounding error. It is the difference between a form that gets completed and one that gets started and abandoned. Baymard’s broader work finds that a quarter of users have abandoned a purchase specifically because the flow was too long or too complicated, and that checkout design improvements alone can lift conversion on large sites.
The number to take from this is not 8. It is the principle underneath it. Every field has a cost, that cost is measurable, and almost every form is carrying fields nobody has justified.
Why field count works the way it does
A form field is not a small ask. Each one requires the visitor to decide three things. Whether they know the answer, whether they are willing to give it, and what you will do with it.
The second and third questions are where B2B forms lose people. A phone number field on a demo request is not experienced as one more box. It is experienced as a decision about whether to accept sales calls. Company size and budget dropdowns are experienced as qualification, and a visitor who suspects they will be disqualified often leaves rather than find out.
This is why “we need that data for lead scoring” is a weak defence of a field. The data has value only if the form gets completed. A qualification field that removes a third of your submissions has to improve lead quality by more than a third to break even, and that is rarely tested.
Which fields to keep
For a B2B enquiry form, the honest minimum is usually three. Name, work email, and what they want. Everything beyond that needs a reason.
Some fields earn their place. Some do not.
- Company name is often redundant. It is derivable from a work email domain, and asking for it duplicates information the visitor has already given.
- Phone number suppresses completion when it is required and adds little when it is optional and unused. If sales does not call within a day, it is not earning its cost.
- Job title as a free text field is nearly useless for routing because people type inconsistent things. As a short dropdown it can be useful, if the options are distinct.
- Budget and timeline are the two fields most likely to be added for internal convenience and most likely to cost submissions. If the sales conversation establishes both in the first five minutes, the form does not need them.
- Message is worth keeping. It is the only field where the visitor tells you something you could not infer, and it is optional in effect because people write as much as they want.
The test for any field is simple. If nobody can name the decision that changes based on that answer, remove it.
Labels, errors and the mechanics that cost completions
Field count is the headline finding. The mechanical details underneath it account for a large share of real-world abandonment, and they are cheaper to fix.
Labels belong above fields, and they belong there permanently. Placeholder text used as a label disappears the moment someone starts typing, which means anyone who is interrupted has to clear the field to remember what it wanted. It also fails for people using screen readers.
Errors belong next to the field that caused them. A message at the top of the form saying “please correct the errors below” makes the visitor hunt. Baymard’s testing repeatedly finds that error handling, not field count, is where people give up on forms they had every intention of completing.
Errors should appear after the field is left, not while it is being typed. Validating an email address on the third keystroke tells someone their input is wrong before they have finished giving it.
Error messages should say what to do. “Invalid input” describes the system’s state. “Enter a work email address, for example [email protected]” describes the fix.
Never clear a form on error. Losing typed input is the single most reliable way to lose a submission entirely.
Say what happens next. A line under the button that reads “We reply within one working day” does more for completion than most design changes, because the main hesitation before submitting is uncertainty about what you have just signed up for.
The average checkout asked for 11.3 form fields in 2024. Testing finds most sites need eight.
Baymard Institute
Does splitting the form into steps help
The usual response to a long form is to break it across several screens. Sometimes this works. Often it hides the problem rather than solving it.
Multi-step forms help when the steps are different kinds of question, when there is a visible indication of progress, and when the first step is the easiest one. Starting with a single low-effort question and revealing the rest afterwards works because of commitment. Someone who has answered one question is more likely to answer the second.
Multi-step forms do not help when they are the same fields distributed across three screens. The total effort is unchanged, the visitor now cannot see how much is left, and any error sends them backwards. A four-step form asking eleven questions is worse than a one-step form asking six.
The honest test is whether you would be willing to show all the fields at once. If the answer is no, because it would look like too much, the form is too long. Splitting it is a way of concealing that from the visitor, and visitors work it out by the second screen.
What changes for enterprise procurement
Enterprise B2B has a case where longer forms are defensible, and it is worth separating from the general rule.
When a form is the entry point to a formal procurement process, the buyer often expects to supply detail. A vendor registration form, a security questionnaire response, or an RFP submission is not a lead capture form. The visitor arrived intending to complete a process, and brevity is not what they are optimising for.
The distinction is intent. A first-touch enquiry form is talking to someone deciding whether to start a conversation. A procurement form is talking to someone who has already decided. Applying the eight-field rule to the second is as wrong as applying procurement-length forms to the first.
The mistake most B2B sites make is having only one form, sized for the wrong one of these two situations. Usually it is a first-touch form carrying procurement-weight questions, which serves neither.
What this looks like applied
The pattern that works for B2B enquiry forms is short, honest, and specific about the next step.
Ask for the minimum that lets a human respond usefully. Put the qualification questions in the conversation, where a person can explain why they are being asked. Label every field. Handle errors at the field. State the response time and then meet it.
Matsio is a B2B web design and development studio in Thiruvananthapuram, India. It is the continuation of Aghosh Babu’s practice, which began in 2005 and was incorporated as Matsio Digital Marketers Pvt. Ltd. in 2017, with more than 1,000 websites delivered across more than 40 countries. Form simplification is one of the few changes that reliably moves numbers on a B2B site within weeks rather than quarters, because the visitors were already interested. 85% of the studio’s clients return for further work, measured across every engagement since 2005, and a meaningful share of that repeat work starts with fixing the last step of a journey that was otherwise working.
How to audit your own form this week
You do not need testing software to find most of what is wrong.
Open your main enquiry form and write down every field. Next to each one, write the name of the person who uses that answer and the decision it changes. Delete any field where you cannot fill in both columns.
Then submit the form yourself, deliberately wrong. Leave the email malformed. Skip a required field. Watch what the form does. If it clears your input, puts the error somewhere other than the field, or tells you something is invalid without telling you what valid looks like, you have found the problem that is costing you more than field count.
Finally, submit it correctly and see what happens. If the confirmation is a blank page or a generic thank-you with no indication of when someone will reply, the form is working and the experience around it is not.
Related reading. what an A/B test measures covers how to tell whether a form change worked, and heatmaps on a B2B homepage covers the limits of the tools people usually reach for first. For the wider question of what a buyer has already decided before reaching your form, see how buyers research vendors through AI chat.
The short version
Forms fail for boring reasons. Too many fields, labels that vanish, errors in the wrong place, and no statement of what happens after the click.
Baymard’s number is 8 fields for a checkout. A B2B enquiry form should usually be shorter than that. Every field beyond the minimum needs someone willing to name the decision it informs, and most cannot survive that question.
A small thing and a big thing
One small thing to fix on your website today, and one big thing to learn that gets you more leads.
Your information is
safe with us. No spam.