Where do buyers quit the registration form?

We looked at 30 days of Hotjar data on funeralprints.com and tracked, step by step, where people abandon the registration form when they're forced to create an account at checkout. The answer changes what we should test next.

Source: JC's Hotjar funnel (last 30 days) Segment: sessions that visited /login?returnUrl=/checkout — i.e. people shopping who got forced to register

The headline

67% of buyers forced to register at checkout leave before typing anything. Only 15.0% finish the form.

That's 100 completions out of 668 attempts in 30 days. The form itself isn't the problem — the requirement to create an account is.

Step-by-step drop-off

Step	Sessions	% dropped	What's happening
0. Landed on register form	668	—	They see an account is required
1. Clicked First Name	218	67%	Most walk away before typing anything
2. Clicked Last Name	185	15%
3. Clicked Email	145	22%	Likely duplicate-email error or Confirm-Email friction
4. Clicked Confirm Email	127	12%
5. Clicked State	121	5%
6. Clicked Phone	117	3%
7. Clicked Password	102	13%	Password rules, retyping
8. Clicked Confirm Password	100	2%
9. Clicked Terms checkbox	100	0%
10. Clicked Register button	100	0%	If they get here, they finish

Three things we learn

1. Account creation is the main barrier, not the form length

67% of people bail before typing into a single field. They see "create an account" and leave. Removing fields won't fix that — only removing the requirement to make an account will. The real opportunity is guest checkout, especially for Video Tributes (digital products don't need shipping info).

2. Inside the form, Email and Password cause the biggest drops

Email step loses 22% (likely duplicate-email errors, Confirm-Email friction), and Password loses 13%. Name fields lose less than 15% each. So shortening the form by removing names, address, and phone (which is what the PreProd test does) saves roughly 20–25% cumulative — meaningful but not huge.

3. If someone gets past Password, they always finish

Terms checkbox and Register button show 0% drop-off. The finish line isn't the problem. The journey up to Password is.

What this means for the shortened-form test

JC's PreProd site reduces registration from 15 fields to 7 ( see side-by-side ). Based on the drop-off pattern above:

It won't move the 67% who bail before touching the form — they're walking on account creation, not field count.
It should help the 15–22% who drop during Name/Email steps, because shorter forms feel less intimidating and there are fewer spots to stumble.
Realistic upper bound: lift completion from 15% to maybe 20–25% on this segment.

Still worth shipping — small lift, low effort. But the bigger lever is a separate question: can we offer guest checkout for digital products like Video Tributes? That side-steps the account-creation wall entirely for the buyer who just wants a video.

Caveats on these numbers

Hotjar only tracks clicks, not typing. Mobile users who tab or autofill through fields are undercounted at each in-form step. The 67% "walked before clicking First Name" is still reliable — everyone has to do something to start the form.
PostHog can't help yet. The GTM init uses person_profiles: "identified_only", so first-time visitors (the people actually registering) don't get captured. That one-line GTM fix would give us mobile-accurate numbers.
Hotjar samples at "Medium traffic coverage" — it sees roughly 50–60% of sessions, not 100%. Ratios are fine; absolute counts are ~2× higher than shown.

Next steps

Ship the 7-field form when the PreProd QA is done — expected modest lift.
Open a separate card: guest checkout for Video Tributes. Digital product, no shipping needed, no reason to force account creation.
Flip person_profiles to always in GTM so we have PostHog numbers to measure the lift properly after launch.