The pilot with no way to fail
The artifact
Doc: Onboarding Pilot — Guided Setup Owner: Marcus Aliyev, Product Reviewers: Engineering (approved), Design (approved), Support (approved) Status: Approved, kickoff Monday
Background
New-account onboarding has been a known problem for three quarters. Users sign up, land on an empty dashboard, and a meaningful share never configure anything. We have discussed this in three separate planning cycles without shipping a fix.
Guided Setup is a five-step wizard that walks a new account through connecting a data source, inviting a colleague, and creating their first saved view. Design is complete and engineering estimates six weeks.
Objective
Improve the new-account onboarding experience and increase early engagement.
Approach
We will build Guided Setup over six weeks and roll it out to 200 new accounts. During the pilot we will gather qualitative and quantitative insight into how users experience the new flow, iterate based on what we learn, and use the findings to inform the full rollout decision.
Scope
- Five-step wizard (data source, invite, first view, notification prefs, tour)
- Progress indicator and skip option at each step
- Contextual help content for each step
- Analytics instrumentation for step-level drop-off
Timeline
| Week | Activity |
|---|---|
| 1–2 | Build steps 1–3 |
| 3–4 | Build steps 4–5, help content |
| 5 | Instrumentation, QA |
| 6 | Rollout to pilot cohort |
| 7–8 | Monitor, gather feedback |
| 9 | Readout and rollout recommendation |
Success criteria
- Positive qualitative feedback from pilot users
- Improved step-level completion through the wizard
- Increased early engagement in the pilot cohort
- Learnings documented and shared with the wider team
Measurement
We will track wizard step completion, time-to-first-view, and 7-day engagement. We will also run a short in-app survey at the end of the wizard asking users to rate the experience.
Risks
| Risk | Mitigation |
|---|---|
| Six-week build may slip | Weekly checkpoint with engineering |
| Pilot cohort too small to be meaningful | Extend the pilot window if needed |
| Users skip the wizard | Track skip rate; iterate on copy |
| Support load during pilot | Support briefed in week 5 |
Open questions
- Should the wizard be skippable? (Design says yes, Sales says no)
- Do we include self-serve trials in the pilot cohort, or paid accounts only?
Your task
You are Marcus's manager and this lands in your inbox the Friday before kickoff. Redline it: mark up the specific changes that would let this pilot return a negative result, and say what each change is for. Rewrite the lines that need rewriting rather than describing them in general terms. Aim for 300–500 words.
150 more characters before you can submit