Onboarding checklist UX: how to design one merchants finish

In short: An onboarding checklist helps when the product needs more than one merchant action before it pays off, and it hurts when it stands between the merchant and the first result. Keep it to three to five steps, order them by the value each step unlocks rather than by the order your system needs them in, let merchants dismiss it, and measure whether finishing it predicts retention. Most merchants will not complete it. Userpilot's 2025 benchmark puts average checklist completion at 19.2% across 188 B2B SaaS companies, with a median of 10.1%, so the design has to work for the ones who stop at step two.

When does an onboarding checklist help?

A checklist is the right pattern when three things are true. The product needs several independent actions from the merchant before it produces its full result. Some of those actions wait on something outside the app, such as an account approval or a creator replying. And the merchant will come back across several sessions to do them.

An affiliate app fits. The merchant has to create a program, put the signup page somewhere customers will see it, invite the first affiliates, and set up a payout method before the first payout, and those happen over a week or two. A creator outreach tool for TikTok Shop fits for the same reason. An upsell app mostly does not. Pick a product, publish an offer, done.

Shopify's onboarding guidance describes the pattern as a setup guide that "acts as a quick start, with discrete steps that are automatically marked as complete," with a progress indicator, and says that if it is not essential it should be dismissible. The setup guide composition in Shopify's App Home docs is "an interactive checklist with visual progress tracking," meant for the homepage or a dedicated onboarding page.

The checklist is the wrong pattern when it delays the first result. If the merchant could get a visible result from one action, that action belongs on the first screen as a single button, and the checklist starts after it. I cover that first screen in the guide on Shopify app onboarding. The checklist is also the wrong pattern when it is really a feature tour. "Explore the analytics page" is not a step. Nothing happens in the merchant's store when they do it.

How many steps should it have?

Three to five, and the first one should already be done when the merchant first sees it.

Shopify's guidance puts the ceiling at five: "Avoid more than five steps in your onboarding process, as additional steps can lead to merchant drop off." The completion data argues for staying well under it. Userpilot's 2025 benchmark report lists an average checklist completion rate of 19.2% and a median of 10.1%, from 188 B2B SaaS companies using Userpilot's own product. That is first-party vendor data, and each company defined its own checklist, so treat it as a rough signal. The signal is clear enough: four in five accounts do not finish.

Each step should be one action with one visible outcome in the merchant's store. Here is the test applied to a typical affiliate app checklist:

Step as usually written	What is wrong with it	Rewritten
Set up your brand settings	No outcome in the store	Dropped, or moved to settings with defaults
Explore the dashboard	A tour, not an action	Dropped
Create your program	Fine, but it is the first-screen action	Shown as already complete
Add the signup page to your store	One action, one outcome	Put your signup link where customers will see it
Invite affiliates	Vague	Invite your 50 best repeat customers
Configure payouts	Needed, but not yet	Choose how affiliates get paid, before the first payout is due

Five steps, the first already ticked, each one leaving something changed in the store.

Should the checklist be ordered by value or by setup order?

By value. This is the decision that most separates a checklist merchants finish from one they dismiss.

Setup order is the order your system needs things in. Brand settings before the signup page, because the page uses the brand colours. Payout method before invites, because the database wants it. Most first-draft checklists are in this order, because it is the order the code was written in.

Value order is the order in which each step pays the merchant back. The step that produces the first affiliate comes before the step that decides how the affiliate will eventually be paid. The step that gets the first creator reply comes before the step that sets up sample approvals.

Setup order	Value order
1. Brand settings	1. Create your program (done)
2. Commission rules	2. Put your signup link on your thank-you page
3. Payout method	3. Invite your repeat customers
4. Signup page	4. Choose a payout method
5. Invite affiliates	5. Match the signup page to your brand

When a step genuinely must precede another, fold the dependency into the step that needs it, with a default. The signup page needs brand colours, so the step creates the page with the store's name and a neutral style, and the branding step comes later as an upgrade. The invite needs a commission rate, so the program ships with a default rate the merchant can change. Defaults are how you break setup order without breaking the system.

When I design one of these I write each step as a sentence that ends with what the merchant now has. "Put your signup link on your thank-you page, so customers who just bought can join." If the sentence has no second half, the step is setup, not value.

What does the research say about progress?

The relevant work is Kivetz, Urminsky, and Zheng's 2006 paper in the Journal of Marketing Research, The Goal-Gradient Hypothesis Resurrected. In a real café reward program, customers bought coffee "more frequently the closer they are to earning a free coffee." The experiment most quoted in product circles is the second one: customers given a 12-stamp card with two "bonus" stamps already on it completed the ten required purchases faster than customers given a regular 10-stamp card. The authors also found that a stronger tendency to accelerate toward the goal predicted greater retention in the program.

Two things follow for a checklist, and one warning.

First, show progress. A progress indicator is not decoration, and Shopify's guidance recommends one. Second, start the merchant past zero. Install, OAuth, and the first-screen action are real steps the merchant really took, so the checklist should open with them ticked. That is the honest version of the two bonus stamps.

The warning is that the honest version is the only one worth shipping. A checklist that says "20% complete" before the merchant has done anything is noticed, and once it is noticed the merchant reads every other number in your app with the same suspicion. Pre-complete what was done. Do not pre-complete what was not.

What should happen when the merchant skips it?

They will, and the design should assume it.

• Make it dismissible, as Shopify's guidance says, and keep it reachable. A collapsed card on the home page or a "Setup" item in the navigation. Never a modal that reappears.
• Move the skipped steps into context. The step "Put your signup link on your thank-you page" belongs in the empty state of the affiliates list, because that is the screen a merchant with no affiliates will be staring at. The guide on empty state design covers that screen. A checklist that is skipped and an empty state that repeats its step are two routes to the same action, and the merchant only needs to find one.
• Trigger reminders from the merchant's state, not from the calendar. "You have three affiliates and no payout method, and the first payout is due in six days" is a message a merchant will act on. "Finish setting up your account" is not.
• Route the stores that matter to a person. A high-volume store that has dismissed the checklist at step two is worth a message from customer success. This is the list a CS team should have on screen each morning.

If most merchants skip the checklist in the first session, that is usually not a checklist problem. The first screen asked for too much, and the checklist is where they stopped. Fix the first screen.

How do you measure completion?

Build it as a funnel in your product analytics tool, aggregated by store rather than by user, with the install as the first step and each checklist item as a step. Then read five things:

1. Reach rate per step. The share of installs that completed each step, read as a funnel, which shows you exactly which step the drop happens at.
2. Overall completion rate, compared with the 10 to 19% range from Userpilot's data as a sanity check and then against your own trend.
3. Time from install to each step, as a median. A step that takes a median of four days is waiting on something, and the message for that step should say what.
4. Dismiss rate, and the step at which merchants dismiss.
5. The one that decides the design: 90-day retention by number of steps completed. If merchants who completed step three retain no better than merchants who stopped at step two, step three is not earning its place. Cut it or move it. The guide on SaaS activation rate covers the same test for choosing the activation event, and it is the same query.

That fifth number is also the honest answer to "what is a good completion rate." A checklist with 15% completion where steps one and two carry the retention lift is doing its job. One with 40% completion where no step predicts retention is a well-designed distraction.

Where to start

1. Write out your current checklist and, next to each step, what has changed in the merchant's store once it is done. Drop any step with nothing in that column.
2. Reorder the remaining steps by how soon each pays the merchant back, and fold any true dependency into the step that needs it, with a default.
3. Tick the steps the merchant has genuinely done by the time they first see the list. Do not tick anything else.
4. Put each step's action into the empty state of the screen it would fill, so a merchant who dismissed the checklist still finds it.
5. Build the per-step funnel by store and add 90-day retention by steps completed. Read it monthly and cut the steps that do not move it.