#UX & Conversion

A contact form should not require a packed lunch

Zulu Shack illustration for “A contact form should not require a packed lunch”: packed lunchbox, set aside

The first question in a contact-form review should be what happens after someone presses Send.

If the business uses the submission to arrange a conversation, the form needs enough information to start that conversation. It probably doesn’t need to complete the entire discovery process before anyone has said hello.

A fictional web studio might ask for a name, contact route and a short description of the project. Its existing form also requires annual revenue, a precise page count, a platform choice and a launch date.

Some of those details may become useful. Asking for them now can turn a simple inquiry into an assignment with research requirements. We wanted to hear about the project, and somehow we’ve assigned a small dissertation.

Nielsen Norman Group’s March 2025 EAS framework offers a useful way to examine that effort: eliminate unnecessary work, automate appropriate parts and simplify what remains.1 The point is to make the task easier, rather than merely rearranging the same questions.

I like working backward from the person who receives the inquiry. Which answer changes what they do next? Which answer helps them prepare? Which one gets copied into the CRM and then spends the rest of its career untouched?

That discussion can reveal a question the team included because it might someday be interesting. A contact form is an expensive place to maintain a wish list of possible future data.

Slow pages and outdated forms were holding Office Accomplice back. See how the redesign fixed both.

Keep the questions that improve the next step

For our sample studio, a broad service choice could help route the message. A free-text description could reveal what the prospective client is trying to achieve. Requiring them to choose between technologies they may not know adds less value.

Let people say they’re unsure. If a launch date is only a preference, don’t present it as a firm commitment. If a budget question is useful at this stage, explain why and give people an honest way to answer.

Automation needs the same care. A signed-in customer may be able to confirm existing contact details instead of typing them again. An automatically suggested value should remain visible and correctable.

Inferring a company from an email domain is a different matter. It may be a clue, but it isn’t always a verified fact. The form shouldn’t quietly become more certain than the person completing it.

Simplifying also includes labels, instructions and error recovery. Explain a required format before a failed attempt where practical. Preserve answers when a validation error occurs. Make the field and the problem understandable without relying on color alone.

Related read: The form found a problem. Now help the person finish.

Baymard’s checkout research is relevant background on the effort created by fields, though its ecommerce findings don’t establish a specific conversion gain for a contact form.2 Our studio needs to observe its own visitors.

Watch someone complete a realistic inquiry. Notice where they pause, leave to find information or enter a placeholder just to continue. A form can receive every required answer and still collect very little truth.

Then review whether the shorter or clearer version produces usable inquiries for staff. Completion matters, but the handoff matters too.

A good contact form leaves enough room for the person to begin in their own words. The detailed discussion can happen with someone who knows which question to ask next. By then, we may even have earned the right to ask about the page count.

Turn reading into results

Like what you’re reading? Let’s put it to work on your brand.