Why Half Your Patients Quit the Intake Form (and How to Fix It)

You spent real money getting someone to your site. They read the condition page, decided this might solve their problem, and clicked "Get started."

Then they hit a form with 34 fields, no indication of how long it will take, and a request for their insurance details on screen two.

A large share of them close the tab. And unlike an abandoned shopping cart, you often have no way to reach them, because you never got an email address.

Intake is usually the leakiest part of a telehealth funnel, and it's also the most fixable.

The tension nobody admits

Clinical and product want opposite things.

Your medical director wants complete history, thorough screening, and nothing missed. That's their license and the patient's safety. Entirely reasonable.

Product wants the shortest path to a booked consult. Also reasonable.

Both are right, and the resolution is almost never "ask fewer questions." It's asking them in a better order, at better moments, in a better format.

Fix the order first

The highest-leverage change costs nothing: ask the disqualifying questions first.

If someone in a state you're not licensed in can't be treated, find that out on screen one. Not screen seven, after they've typed out their full medical history. You've wasted twelve minutes of their time and generated a partial record you now have to handle.

Then order the rest by effort and sensitivity:

1. Disqualifiers. State, age, the one or two conditions that rule someone out.
2. Easy and low-stakes. Name, email. Now you can reach them if they drop off.
3. The clinical substance. History, symptoms, medications.
4. The sensitive and the tedious. Insurance, ID upload, payment.

Getting an email early is the difference between an abandoned form and a recoverable one.

Make it resumable

People fill out intake on a phone, on a couch, in the middle of an evening. Someone knocks on the door. A kid needs something. The tab gets closed.

If returning means starting over, they don't return.

Save progress on every step, server-side, keyed to the record you created when they gave you an email. Then send one gentle reminder with a resume link — not a marketing sequence, just a way back in.

This single feature recovers more intakes than any amount of copy rewriting.

One question at a time beats a wall

A 34-field form looks like homework. The same 34 fields across a series of short screens with a progress bar feel manageable, because each screen is trivially answerable and progress is visible.

It's the same total work. It doesn't feel like it, and feeling is what determines whether someone finishes.

A few specifics that matter:

• Show real progress. "Step 3 of 5" is honest. A bar that jumps from 20% to 80% is not, and people notice.
• Use the right input for the job. Date pickers for dates, numeric keyboards for numbers, a searchable list for medications. Typing a medication name into a plain text box on a phone is miserable and gives you dirty data.
• Ask conditionally. Don't show pregnancy questions to everyone. Branch based on earlier answers.
• Explain anything invasive. One line next to the ID upload — "we're required to verify identity before a provider can prescribe" — reduces drop-off measurably. People abandon what they don't understand the reason for.

The screens where people actually quit

Instrument this, and you'll almost always find the same four culprits.

The insurance screen. Long, confusing, and people often don't have their card nearby. Consider whether you can capture it after the consult is booked rather than before.

ID upload. Camera permissions, bad lighting, file size errors. Make it forgiving — accept photos from the gallery, handle large files, never lose their progress on a failed upload.

The medication list. If it's a free-text box, it's both painful to fill in and useless to you. Searchable autocomplete solves both.

Payment. If it comes before they've spoken to anyone, expect drop-off. Be clear about what they're paying for and what happens if they're not eligible.

Measure the funnel, not the total

"Intake completion is 52%" doesn't tell you what to fix. Per-step completion does.

Track how many people start each step and how many finish it. The step with the steepest drop is your next piece of work. Fix it, measure again, move to the next one.

Also watch time on step. A step people complete but spend ninety seconds on is confusing them, even though it isn't showing up as abandonment. That's a fix waiting to happen.

Two cautions. Do this with server-side analytics, not a browser session recorder, because a recorder capturing intake fields is capturing PHI. And never put condition or clinical detail in a URL or page title — that leaks straight into your analytics.

What not to do

Don't cut clinically necessary questions to boost a number. If your medical director says it's required, it's required. Move it, format it better, explain it — but don't remove it because it costs you conversions.

Don't fake urgency. Countdown timers on a medical intake are gross and they erode the trust you need for someone to hand you their health history.

Don't hide the length. "About 5 minutes" up front sets expectations. Discovering on step 6 that there are nine more is when people quit.

The one that pays for itself

If you only do one thing: get an email address on the second screen and make the form resumable.

Everything else is optimization. That one turns your biggest silent loss into something you can follow up on.

---

We build patient intake flows, scheduling, and the BAA-covered infrastructure behind them. See how we work with telehealth companies.