# The 5-leak onboarding form audit

Run your current onboarding through these five failure modes. Fix the one
where your clients actually stall — leave the rest alone.

## Leak 1 — Effort with no end in sight

- [ ] Is the form split into steps of 3–5 questions?
- [ ] Is there a visible progress indicator?
- [ ] Do you state the time cost up front ("about 20 minutes")?

**Fix:** multi-step with progress. Single-page walls of fields tell the
client "this will eat your evening," and they're right to close the tab.

## Leak 2 — Hard asks come first

- [ ] Do easy context questions come before uploads and access requests?
- [ ] Can a client skip a hard item and come back without losing progress?

**Fix:** order by effort — easy first, thinking middle, homework last.
People abandon at the first item they can't do immediately.

## Leak 3 — It's actually four forms in three tools

- [ ] Does the client get ONE link, or a form + folder invite + e-sign
      request + invoice?
- [ ] Can you see at a glance which pieces are done?

**Fix:** one link, everything behind it, in order. This is the single
highest-impact change.

## Leak 4 — Progress is invisible

- [ ] Does the client see "3 of 5 steps done"?
- [ ] Is there a real anchor (kickoff date) the form feeds into?

**Fix:** show completion state and tie the form to the kickoff: "we'll use
your answers to run Thursday's call."

## Leak 5 — Nothing happens when they stall

- [ ] Does anything trigger automatically after 48 hours of inactivity?
- [ ] After two automated nudges, do you switch to a personal message/call?

**Fix:** inactivity reminders at ~48h and ~96h with the link and what's
left; then a human conversation, because at that point something else is
going on.

## Measure

- Time-to-complete (median): healthy is under 48 hours for a committed client
- Where clients stall: same step twice = that step is the problem

---

From the Onvoy blog: https://onvoy.app/blog/why-clients-dont-finish-onboarding-forms
Progress tracking and reminders built in: https://onvoy.app
