Free Playbook · End-to-End Beta

Beta Program
/ Playbook.

Recruit, onboard, collect, close-loop. The runbook senior PMs share with new hires.

Beta name: _____
Owner: _____
Start: _____

A bad beta is worse than no beta. It pollutes your data, burns your best customers' patience, and sets the new feature up to look like a flop on launch day. This playbook moves through the four phases — Recruit, Onboard, Collect, Close-Loop — with the rules senior PMs apply at each step.

Beta Program Goals

Three valid goals — pick one as primary:

Validate desirability. Will the target users actually use this? Behavior > opinion.
Validate usability. Can users complete the core task with the current UX, without a sales-engineer hand-hold?
Generate launch artifacts. Quotes, case studies, founding-customer logos, social proof for GA day.

Three goals that will sink your beta — if any of these is on the doc, rewrite the doc:

"Free QA." Beta is not your test environment. Ship working software, then test desirability.
"Hit a revenue target." Beta users are early adopters, not your ICP at scale. Don't extrapolate.
"Build the feature with the customer." That's a custom-dev contract. Pick a name and pricing for it.

Recruitment Criteria + Scorecard

Score every applicant 1–5 on each criterion. Aim for a total of 18+ out of 25 across your final list.

Criterion	What "5" looks like	Score 1–5
ICP fit	Lives at the center of your ICP — segment, role, use case all match.	__/5
Urgency of pain	Has tried 2+ workarounds. Brings up the pain unprompted.	__/5
Willingness to give feedback	Has shipped feedback to other vendors before. Replies in <24h on the call.	__/5
Technical comfort	Can debug their own auth issue, comfortable in a Slack-only support model.	__/5
Public-credit potential	Recognizable name or logo. Active on LinkedIn / X. Will say yes to a case study.	__/5

Recruitment Channels

Warm list — existing customers who've expressed interest. Conversion: 30–50% of replies. Highest-quality, smallest pool.
Waitlist — anyone who signed up for early access. Conversion: 5–15%. Watch for tire-kickers.
Sales pipeline — opps that lost or stalled because of the missing feature. Conversion: 20–30%. High intent.
Community — Slack, Discord, your subreddit. Conversion: 2–5%. Best for low-stakes consumer betas.
Cold — targeted LinkedIn DMs to ICP-perfect names. Conversion: 1–3%. Slow, but the highest-fit accounts you'd never have met otherwise.

Kickoff Communication Template

Send this within 24 hours of acceptance. Replace bracketed placeholders. Personalization beats template polish — reference their company, the pain they mentioned, and why they were picked.

Beta Acceptance Email

Subject:You're in — [Product] [Feature] beta starts [date]

From:[Your name], PM at [Product]

To:[Customer first name]

Hi [first name], Welcome to the [Feature] beta. You're 1 of 8 we picked — chosen because you flagged [specific pain] back in [month], and we want to know whether what we built actually fixes it. Here's what to expect over the next 4 weeks: • Day 0 (today): I'll DM your access link in Slack. Five-minute setup. • Day 3: Quick "did you make it in?" check. • Day 7: 20-min call with me — what's working, what's not. • Day 14 + 21: Same beat. Plus a 5-question survey each week. • Day 28: Wrap-up call. We decide together if you stay on as a founding customer. What we ask in return: • Brutal honesty — "this is great" is the least useful thing you can say. • Reply within 48h to our weekly pulse so we know you're alive. • Permission to quote you (anonymized or named — your call) when we launch. You'll have a direct line to me — [Slack/email] — for the entire beta. No tickets, no queue. Talk soon, [Your name]

Onboarding Sequence

Day 0

Land & Activate

DM with the access link, the 5-min setup video, and a personal sentence about why they were picked. CTA: log in, complete the activation step.

Watch: % logged in within 24h. Goal: >80%.

Day 3

First-Use Check

Personal 1-line message: "Were you able to [core task]? Stuck anywhere?" Don't broadcast — it's individual.

Watch: anyone who hasn't activated. Reach out 1:1 before they fall off.

Day 7

First Pulse + Live Call

5-question survey (clarity, friction, value, blocker, NPS-style intent). 20-min Zoom with the user. CTA: show me how you used it.

Watch: WAU, time-to-first-value, top 3 friction points across calls.

Day 14

Mid-Beta Pivot Check

Based on Days 0–7 signal, ship the highest-leverage fix. Send a short "you said X, here's what we did" note — the close-loop muscle starts here.

Watch: did the fix move usage? Did the complaint stop appearing?

Feedback Collection Cadence

Weekly pulse survey — 5 questions, <90 seconds. Same questions every week so you can see the trend.
Bi-weekly 1:1 call — 20 minutes, screen-share the actual product usage. Record with consent.
Always-on Slack channel — one shared channel for the cohort + you. Never private DMs only — users see each other's feedback and pile on.

Tag every piece of feedback as a signal. In MindBacklog, route the channel to a dedicated "Beta — [feature]" tag so it shows up in the same triage queue as customer support and Slack.

Triage Workflow

Fix immediately — auth bugs, data loss, "I can't log in." If two betas hit it, ship today.
Capture as signal — feature requests, UX papercuts, "would be nice if…" Score and rank weekly. Do not promise.
Defer (with explanation) — out-of-scope asks. Reply: "Not in the beta scope — tracked for v1.1." Saying no graciously builds trust.

Graduation Criteria (Beta → GA)

Don't ship to GA until all three gates clear:

Activation gate. ≥75% of beta users completed the core success event at least once.
Retention gate. ≥50% of beta users return weekly by Week 3 without prompting.
Quality gate. Zero P0/P1 bugs open. <3 P2 bugs. Documented known limitations list.

Public Credit Playbook

Case study. 30-min interview, written write-up, customer review and approval. Your strongest sales asset for the next 12 months.
Social shoutout. Pre-written LinkedIn post with their tag — they share. Always offer text they can paste.
Founding-customer logo wall. Logo on the marketing site for as long as they're a customer. Costs nothing, prints conviction.

Common Beta Failure Modes

Too many betas in flight. Three concurrent betas means none of them gets the PM's attention. Limit: one major beta per PM at a time.
No exit criteria. "We'll launch when it feels ready" stretches forever. Define the gates above — and stick to them.
Feedback theater. Collecting feedback you never act on. If you can't ship a single change inside the beta window, you're running a focus group, not a beta.