About half of the people who start a web form finish it. Zuko's benchmark of 739 forms and 93 million form views puts the average starter-to-completion rate at 51.7% and the average view-to-completion rate at 34.6%, measured between January 2022 and April 2025. Contact forms sit near the bottom: only 14.5% of contact form views end in a submission.
This page collects the form data that is actually published, from Zuko's form analytics, Formstack's Form Conversion Report and HubSpot's research, with the definition and the age of each number next to it. Several of the most quoted form statistics turn out to be a decade old or observational, and we say which.
How many people finish a form
Zuko is a form analytics vendor, so its numbers come from customers who chose to instrument their forms, skewed towards financial services, ecommerce, insurance and gambling. It is still the largest published field-level dataset, and the only one that separates people who looked at a form from people who started it.
Two definitions matter. View to completion is submissions divided by people who saw the form. Starter to completion is submissions divided by people who interacted with at least one field. The second measures friction once someone has committed.
| Form purpose | Sessions tracked | Desktop | Mobile | Overall starter to completion | View to completion |
|---|---|---|---|---|---|
| Registration | 26,582,277 | 63.0% | 57.2% | 60.7% | 46.6% |
| Purchase | 20,179,282 | 56.9% | 51.4% | 54.4% | 40.0% |
| Onboarding | 4,678,212 | 53.2% | 47.8% | 51.8% | Not published |
| Application | 27,575,784 | 53.9% | 45.0% | 51.6% | 37.7% |
| Contact | 321,937 | 51.2% | 39.4% | 47.4% | 14.5% |
| Comparison | 7,583,511 | 48.2% | 44.9% | 46.4% | 28.6% |
| Enquiry | 4,797,087 | 47.7% | 37.7% | 42.7% | Not published |
| All forms | 93,022,997 | 54.5% | 47.5% | 51.7% | 34.6% |
Source: Zuko form type benchmarking and Zuko online form statistics, data January 2022 to April 2025. Device columns are starter to completion.
The contact form row is the one most B2B marketing sites should look at. Nearly half of the people who start a contact form finish it, which looks acceptable, but only 14.5% of the people who see one submit it. Dividing the two Zuko figures suggests roughly three in ten contact form viewers ever start typing. That is our arithmetic on Zuko's averages, not a published figure, and it says the main leak on a contact form is before the first field, not inside it.
How long forms are, and what extra fields cost
Zuko's field counts. Zuko reports the average number of labelled fields by form purpose: 44.2 for comparison forms, 38.9 for applications, 31.5 for purchase, 21.2 for configuration, 19.2 for contact, 17.4 for enquiry and 16.0 for onboarding. These count every labelled input Zuko tracks, so they run higher than the number of questions a visitor would count. Treat them as relative, not as a target.
HubSpot's landing page analysis. Dan Zarrella's study of over 40,000 HubSpot customer landing pages is the source behind most "fewer fields" advice. It found that conversion rates fall as fields are added, but not as steeply as he expected. Adding single-line text fields barely moved conversion. Multiple text areas had a strong depressing effect, and several drop-downs also hurt. A companion post found that asking for age, a phone number or a street address was associated with lower conversion. Neither post publishes percentages, and the analysis is much older than the June 2025 page update.
Popups behave differently. Sleeknote's 2025 analysis of more than 329 million popup impressions shows one-field forms converting at 4.41%, two fields at 2.90%, three at 1.83%, four at 2.27% and five at 3.26%. The curve is not a straight decline, and these are mostly email-capture popups, so the lesson does not transfer cleanly to a demo or quote form.
| Source | Sample | Finding | Type of evidence |
|---|---|---|---|
| HubSpot, Zarrella | 40,000+ landing pages | Conversion falls slightly per field; text areas, drop-downs, age, phone, address hurt more | Observational, no percentages published |
| Sleeknote, 2025 | 329M+ popup impressions | 1 field 4.41%, 3 fields 1.83%, 5 fields 3.26% | Observational, popups only |
| Formstack, 2015 | 650,000+ form users | Survey forms averaged 22 fields and still converted 14% of visitors | Observational, platform data |
| Baymard, 2024 | Ecommerce checkout benchmark | Average checkout has 11.3 fields, down from 12.7 in 2019; most need only 8 | Usability benchmark |
| HubSpot survey, 2019 | 173 marketers | Average form has 4.92 fields; five is the most common answer | Self-reported |
Sources: HubSpot blog, Sleeknote, Formstack Form Conversion Report 2015, Baymard Institute
Baymard's finding is the most useful general rule here: in checkout research, the number of fields affected usability far more than the number of steps. Its 2024 benchmark puts the average checkout at 11.3 fields, down from 11.8 in 2021, while most sites need only 8. On a lead form the equivalent question is which fields sales actually uses in the first call. Our Demo Request Conversion Rates 2026 post covers the counterpoint for high-intent forms: RevenueHero's top performers book about the same share of qualified leads with two fields as with thirteen.
Which fields people abandon
Zuko's field-level research tracked 1,362 forms over 12 months and recorded, for each common field, how long people spent in it, how often they came back to it and how often it was the last field touched before someone left.
| Field | Forms in sample | Mean time in field | Mean abandonment | Median abandonment |
|---|---|---|---|---|
| Password | 238 | 7.2 s | 10.50% | 5.95% |
| 1,003 | 6.5 s | 6.41% | 2.98% | |
| Phone | 1,028 | 5.0 s | 6.28% | 3.29% |
| Name | 839 | 3.5 s | 5.27% | 2.52% |
| Postcode | 505 | 3.8 s | 4.82% | 1.57% |
| Address | 301 | 7.4 s | 4.32% | 1.70% |
Source: Zuko, which form fields cause the biggest UX problems, page updated 2026-10-08
Two things stand out. Medians are roughly half the means, so a few badly built forms drag the averages up; a typical phone field loses about 3% of the people who reach it, not 6%. And phone has the highest median of the fields a B2B lead form usually asks for, which matches HubSpot's older finding that asking for a phone number was associated with lower conversion. Password sits higher still, which is relevant to any free-trial or signup form.
Formstack's 2015 report made a similar point without numbers: some of the most used fields, text, name and email, were also the most common places people abandoned, and radio buttons were a leading cause of high abandonment.
Do multi-step forms convert better
The headline number comes from Formstack's 2015 Form Conversion Report: across more than 650,000 users of mobile-responsive Formstack forms, multi-page forms converted 13.85% of visitors and single-page forms 4.53%. The report itself explains the likely cause. Longer forms were the ones that used page breaks, and they tended to carry progress bars and more generous spacing. It compares different forms built by different customers for different purposes, so it shows an association, not that splitting a form into steps triples its conversion.
| Source | Design | Result | Caveat |
|---|---|---|---|
| Instapage | A/B test on its own lead form, 25,500+ visits per variation | Multi-step won by 18% in the summary, 21.4% in the body | Vendor test, year and confidence not stated |
| Sleeknote, 2025 | 329M+ popup impressions | Single-step 3.81%, multistep 3.75% | Popups, mostly email capture |
| HubSpot survey, 2019 | 173 marketers | Multi-step users self-report 29.76% conversion versus 16.05% | Self-reported, small sample |
| Formstack, 2015 | 650,000+ form users | Multi-page 13.85%, single-page 4.53% | Different forms compared |
| Venture Harbour | Own consulting enquiry form | 0.96% to 8.1% after replacing a basic contact form | Before and after, no traffic stated |
Sources: Instapage, Sleeknote, HubSpot, Formstack, Venture Harbour
The pattern is consistent once the weak evidence is set aside. When a long form is split into steps and the first step asks something easy, completion tends to improve, by something closer to 20% than 200% in the one controlled test with stated traffic. When the form is already one or two fields, steps add nothing. Venture Harbour's 743% figure, which it quotes as "743% increase in leads", replaced a basic WordPress contact form, so the baseline was weak, and no traffic or test length is given. Treat it as a single case on the author's own site, not a benchmark.
HubSpot's own test points the other way on layout. On its demo landing page, a two-column form converted 22% better than a one-column form with the same fields, significant at 99% confidence. One test does not overturn single-column advice, but layout rules need testing on the page in question.
The mobile gap
Zuko's sessions are 73.8% mobile, so this is not a dataset where mobile is a minority. Starter-to-completion averages 54.5% on desktop and 47.5% on mobile, a gap of 6.95 percentage points. On a view-to-completion basis the industry page gives 37.2% on desktop and 31.3% on mobile. Desktop converted higher than mobile in every industry except gambling.
The gap is not uniform by form type. Working from Zuko's table, it is 11.8 points on contact forms and 10.0 points on enquiry forms, against 3.3 points on comparison forms and 5.5 on purchase forms. Contact and enquiry forms are the two types B2B sites rely on most, and they lose the most on phones.
Autofill closes part of it. Zuko reports 62.2% completion when autofill was triggered against 30.2% when it was not. Google's Chrome team found users abandoned address and payment forms 75% less often with autofill and spent about 35% less time filling them, across thousands of forms on high-traffic US sites. Both studies are correlational: people who use autofill may simply be more practised. Still, autofill only works when fields carry correct names and autocomplete attributes.
Labels are often broken. WebAIM's February 2026 scan of the top million home pages found an average of 6.9 form inputs per page, with 33.1% of them not properly labelled, and missing labels on 51% of pages. An unlabelled field is harder to fill on a phone, harder for a screen reader and less likely to autofill.
What Formstack's form-type numbers add
Formstack's 2015 report is old, but it is one of the few public sources that reports conversion as submissions over all form visitors by form type: contest forms 35%, surveys 14%, event registration 11%, lead generation 11%, order and payment 9%, donation 7% and contact forms 1%. The contact form figure, down from 3% in its 2014 report, sits in the same territory as Zuko's 14.5% view-to-completion once you allow for the different samples and a decade between them. Both say the generic contact form is the weakest form on most sites.
What this means for your lead forms
Measure the form before redesigning it. Track views, starts and completions separately, because a contact form that converts 14.5% of views and 47% of starters needs a better reason to begin, not fewer fields. Remove the fields sales does not use in the first conversation, and be most careful with phone, text areas and drop-downs. Split a form into steps only when it is genuinely long, and put the easiest question first. Then check the mobile version separately, with labels and autocomplete attributes in place, since that is where contact and enquiry forms lose ten points or more.
On the marketing sites we build in Webflow, we would set up view, start and completion tracking on every lead form before launch, so the first redesign decision rests on the site's own numbers rather than these averages. Our landing page service page explains how we scope that work.
