When lead volume drops, the first instinct is to rewrite something. New headline, new button text, maybe move the form higher on the page.

Meanwhile the form itself is throwing a validation error on iPhones that nobody has seen, or routing submissions to an address that stopped being monitored when someone left in March. Those failures do not show up in a heatmap and they do not announce themselves. The page looks fine. The form looks fine. It has been quietly eating leads for six weeks.

This is an audit, not an essay. Work down it and check each item on your own site.

First, ignore the number everyone quotes

You have seen the claim that roughly 70% of forms are abandoned. That figure comes from Baymard Institute's aggregation of around 50 studies on ecommerce cart abandonment, and Baymard is careful to say that a large share of those abandonments are natural browsing behavior, people comparison shopping or saving items for later.

A cart addition and a quote request are not the same act. Applying a checkout benchmark to a B2B contact form tells you nothing useful, and the field-count rules built on top of it are weaker than they sound. Fewer fields is generally better. It is also the least valuable thing on this list, because a five-field form that delivers reliably beats a three-field form that does not deliver at all.

Fix delivery first. Then argue about fields.

The form audit

Part one: does the form work

Submit it from your phone, on cellular data. Not on office wifi, not in a desktop browser resized to look like a phone. Real device, real network. This single test catches more than the rest of the list combined.

Check that validation errors are visible. Tap submit with a field deliberately wrong. If the error message appears above the fold on desktop but below the keyboard on mobile, the visitor taps submit, sees nothing happen, and leaves believing your site is broken.

Confirm the keyboard matches the field. Phone fields should trigger the number pad, email fields the email keyboard. Wrong keyboard types add friction on every single mobile submission.

Tap every checkbox and radio button with a thumb. Touch targets that work with a mouse are frequently too small for the average thumb, and a mis-tap on a required consent box produces an error the visitor does not understand.

Test autofill. If your input names follow standard conventions, browsers fill name, email, and phone automatically. If they are named field_1 and field_2, nothing autofills and every visitor types everything.

Count the required fields, then justify each one out loud. Anything you cannot defend as necessary before the first conversation should be optional or gone. Sales can ask on the call.

Part two: does the submission arrive

This is the half almost nobody audits and where the expensive failures live.

Send a real test lead and confirm it lands everywhere it should. Inbox, CRM, and any notification channel. Check all three, not just the one you happen to be watching.

Check where the notification goes. A distribution list, a shared inbox, or a person who no longer works there all look identical in the form settings. Only one of them results in a reply, and the difference between them decides how long the lead waits before a human sees it.

Look inside your spam and junk folders. Notifications from your own domain, sent by a form plugin through a third-party mail service, land in spam more often than anyone expects. Also check whether the form's own spam filtering is quarantining real submissions from free email domains.

Verify the CRM integration fires on submit, not on a schedule. Batch syncs that run hourly are fine for reporting and terrible for response time.

Part three: can you see it in the data

Confirm the thank-you page or confirmation event actually fires. If your conversion tracking depends on a redirect that no longer happens, your reporting will show a drop that your inbox does not.

Make sure every form is tracked separately. One conversion event covering six forms tells you submissions fell without telling you which form broke.

Watch impressions on the page against submissions. A steady traffic line under a falling submission line is the clearest signal that the problem is the form and not the traffic, which is where the wider diagnostic of where inbound breaks starts.

What we usually find

Three patterns come up over and over in audits.

The form was fine when it launched and broke during an unrelated change. A theme update, a plugin update, a new required field added by someone in sales. Nothing about the change looked risky and nobody retested the form afterward.

The notification is going somewhere real but unwatched. Often an address like info@ that three people technically have access to and none of them treat as theirs. That one is a delivery problem on paper and an ownership problem in practice.

The tracking broke but the form did not. Leads are arriving and being answered while the dashboard reports a collapse, which sends everyone off optimizing a page that was never the problem. Worth ruling out before any redesign, and worth reading alongside the other conversion failures that survive good traffic.

Make it a monthly habit

The audit above takes an hour the first time. Keeping it working takes five minutes a month.

Submit one real test lead from a phone, on cellular, using an address you check. Confirm it arrives in the inbox and the CRM. Confirm the tracking event fired. Log the date somewhere visible so a gap is obvious.

Do it after every site change, every plugin update, and every time someone leaves the company. Those three events cause most of the breakages we find, and a five-minute check catches them before a month of leads disappears.

If the audit turns up something structural, a form that cannot be fixed with settings or a site where forms break every time anything changes, that is a build problem rather than a marketing one. Our web development team builds and fixes forms as part of the site itself rather than as a plugin bolted on afterward.