Your Onboarding Checklist Is the Leak: A P0/P1/P2 Audit of a 6-Step Trial

Most onboarding teardowns start at the welcome screen. That's rarely where the money leaks.

The leak is usually the checklist you added because something told you new users need "guidance." Six items, a progress bar, confetti at 100%. It feels like help. In practice it's a vendor to-do list wearing a UX hat, and it sits between your user and the only thing that makes them pay.

Below is a full P0/P1/P2 audit of one realistic version of that flow: a B2B scheduling tool called Cadence (fictional, but you've seen it). Sales teams book demos with prospects. The trial is 14 days, no credit card.

The flow, as the user experiences it

Signup. Email, password, company name. Lands on an empty dashboard.
Setup modal fires immediately. "Get set up with Cadence — 0 of 6 complete." Centered modal, dark overlay, small X in the corner.
Item 1: Connect Google Calendar. OAuth screen.
Item 2: Install the Chrome extension. A separate install flow, a browser permission prompt, and a "come back to this tab" instruction.
Item 3: Invite teammates. Email field, role dropdown, send button.
Item 4: Create your booking link. Slug field, timezone, meeting length.
Item 5: Set your availability. A calendar grid the user has to interpret with no context.
Item 6: Watch the 4-minute product tour. Embedded video.
Day 14. Trial expires, upgrade prompt appears. The user has a booking link they never shared and a calendar connection they never used.

Every step here is defensible on its own. Stacked in that order, they form a funnel where the payoff — a link you can paste to a prospect — is item four, sitting behind three chores.

What the checklist is actually optimizing for

Ask one question of every onboarding step: does this serve the user's first outcome, or our setup requirements?

Cadence's checklist is a vendor list. The Chrome extension is a retention mechanism. The teammate invite is a distribution play. The demo video is a support-cost reduction. None of those are bad goals. All of them are being charged to the user's first ten minutes, which is the most expensive real estate you own.

The heuristic mapping gets ugly fast once you lay it out:

Clarity: "Set your availability" assumes the user already knows what availability means inside your product. They don't yet.
Guidance: the checklist guides people through your setup, not toward their result.
Feedback: connecting a calendar produces a spinner and a checkbox. Nothing tells the user what changed or why it matters.
Friction: the extension install is a context switch out of your product and into the browser's permission model — on step two.
Hierarchy: six items of identical visual weight, ordered by engineering convenience instead of value.
Trust: a calendar permission request on step one, before the product has proven anything, reads as a data grab.

The P0/P1/P2 audit

P0 — blocks activation. Fix this week.

The checklist gates the empty state. Until a booking link exists, the dashboard shows nothing useful. Pre-generate a default booking link at signup and put it on the dashboard, above the checklist, with a copy button. The first thing a new user sees should be the thing they came for.
The value step is fourth. Reorder to: create booking link → connect calendar → everything else. Two steps to first value, not four.
Dismissing the checklist removes it permanently. That X kills the only navigation to setup. Someone who closes it at 2 of 6 on day one and returns on day nine has no path back except a support ticket. Pin a one-line progress entry point to the sidebar.
The calendar OAuth has no trust copy. One line above the button — "Cadence reads your free/busy times only. It never reads event titles or attendees." — removes the hesitation that makes people bounce at step one.

P1 — costs you a slice of trials every week. Fix this month.

"0 of 6 complete" frames six as the goal. It isn't. Nobody gets value from 6 of 6. Recast progress around the outcome, and make the remaining items visibly optional.
Copy describes actions, not outcomes. "Connect your calendar" is an instruction. "So Cadence knows when you're free" is a reason. Every item gets a reason.
No skip affordance. Forced linear setup is a friction multiplier. Let people do the two steps that matter and leave the rest for later.
The extension is requested at the wrong moment. Ask when the user actually needs a notification or an overlay — in context, at the moment of need. Install rate goes up and drop-off goes down.

P2 — polish, only after P0 and P1 ship.

Confetti and celebration states. They don't move activation; they make the last 10% feel nice.
The product tour video. Move it to a help link inside the relevant step instead of a checklist item.
Checklist animation, illustrations, and empty-state art.

Before/after: rewriting the checklist so it sells the next step

The original:

Get set up with Cadence ☐ Connect your calendar ☐ Install the Chrome extension ☐ Invite your teammates ☐ Create your booking link ☐ Set your availability ☐ Watch the 4-minute product tour 0 of 6 complete · [Continue setup →]

The rewrite:

Your booking link is 40 seconds away ☐ Create your booking link — we'll use your Google Calendar to find open slots. Nothing to configure yet. [Create my link →]

Optional, whenever you want: add a teammate · install the browser extension · watch the 2-minute tour

What changed and why:

The header names the outcome and the cost. "Your booking link is 40 seconds away" answers "what