lockstep
3 min readNaman Jain

Product launch checklist for software teams

A launch checklist for software teams: engineering readiness based on Google SRE's launch checklist, plus product, support, comms, and post-launch steps.

Launch day. The feature works. Then support learns about it from a customer, the pricing page still shows the old plan, and a traffic spike nobody load-tested slows the whole app. Each was someone's job. None was on a list.

This checklist covers engineering and go-to-market readiness. The engineering half follows the areas in Google's Launch Coordination Checklist from the SRE book, checked on 16 November 2026, which the book presents as an abridged version of Google's original from around 2005.

Engineering readiness

## Architecture
- [ ] Architecture sketch reviewed; new dependencies known
- [ ] New domains or DNS changes ready

## Capacity and performance
- [ ] Traffic estimate, including the launch spike
- [ ] Load test and end-to-end test passed
- [ ] Impact on other critical services checked

## Reliability and failover
- [ ] What happens if a server, zone, or dependency fails
- [ ] Timeouts, retries, and rate limits set
- [ ] Backup and restore tested

## Monitoring
- [ ] Dashboards and alerts for the new feature
- [ ] On-call knows about the launch; runbook linked

## Security
- [ ] Security review done for auth, data, and new endpoints
- [ ] Access control correct before launch

## Rollout
- [ ] Feature flag in place; staged rollout plan
- [ ] Rollback plan tested
- [ ] Launch time chosen; not before a weekend or holiday

## External dependencies
- [ ] Third-party limits checked; graceful degradation if they fail

The SRE checklist's areas: architecture; machines and datacenters; volume, capacity, and performance; reliability and failover; monitoring; security; automation and manual tasks; growth; external dependencies; and schedule and rollout planning, including hard deadlines and whether it's a Monday or Friday.

Product and go-to-market readiness

## Product
- [ ] Acceptance criteria met; definition of done met
- [ ] Analytics events in place for success metrics
- [ ] Pricing, plans, and permissions correct

## Customers
- [ ] Help docs and in-app guidance updated
- [ ] Support trained; macros and FAQs ready
- [ ] Release notes written

## Communication
- [ ] Internal announcement: what, who, when
- [ ] External announcement scheduled: blog, email, changelog
- [ ] Sales and customer success briefed

## Decision
- [ ] Go/no-go owner named
- [ ] Known risks listed and accepted

Running the launch

  1. Assign an owner per item. A checklist item with no name doesn't get done.
  2. Hold a short go/no-go, 24 hours before. The owner decides; record the decision and accepted risks.
  3. Roll out in stages behind a feature flag, watching dashboards at each step.
  4. Staff the launch window, including someone who can roll back.
  5. Review within one to two weeks against the success metrics you set beforehand.

If something breaks, follow your incident response plan and write a blameless postmortem.

Launch decisions

A launch carries a stack of decisions: ship without dark mode, limit to US customers first, keep the old flow for enterprise until March. Each one needs to reach the engineers, support, and AI coding agents working on the feature after launch day.

Where Lockstep fits

Lockstep captures launch decisions from PRDs, Slack, and pull requests, routes each to its owner for approval, and briefs the people and Claude Code sessions working on the feature, so post-launch work follows what was decided.

See Lockstep for product teams.

Sources checked 16 November 2026

For announcing it, see the release notes template. For what done means, see definition of done vs acceptance criteria.

Frequently asked questions

What should a product launch checklist include?
Engineering readiness (capacity, failure handling, monitoring, security, rollout, rollback), product readiness (acceptance criteria met, feature flags set), customer readiness (docs, support training, release notes), communication (announcement and internal briefing), and post-launch review with success metrics.
What is Google's launch coordination checklist?
A checklist from Google SRE, published in the SRE book as an abridged version of Google's original from around 2005. It covers architecture, machines and datacenters, capacity and performance, reliability and failover, monitoring, security, automation, growth, external dependencies, and rollout planning.
Should we launch on a Friday?
Google's SRE checklist lists Mondays or Fridays among the schedule considerations. If you launch late in the week, make sure people who can respond are available over the weekend.
What is a go/no-go decision?
An explicit decision, made by a named owner shortly before launch, that the checklist is complete enough to ship. Record who decided and any known risks accepted, so the call isn't relitigated later.
What should happen after a launch?
Monitor closely during and after rollout, compare results to the success metrics you set beforehand, collect issues, and hold a short review within a week or two. If something went wrong, run a blameless postmortem.

Keep reading