How Do You Conduct A Post Mortem Review?

Q: What’s the biggest mistake teams make?

Turning it into a blame session. Honestly, this kills psychological safety faster than anything. Keep the focus on what happened, not who messed up. The goal is learning, not finger-pointing.

Q: What if the project was a total disaster?

Run the post-mortem anyway—but adjust your approach. Maybe swap in anonymous feedback first. Or bring in that third-party facilitator we talked about earlier. The goal isn’t to relive the trauma; it’s to figure out what to do differently next time.

Q: Can you do this remotely?

Absolutely—just tweak the format a bit. Use breakout rooms for the “Keep / Stop / Try” bucketing. Try a Slack thread for the written-only retro. The key is keeping it interactive, even from a distance.

Q: How do you measure success?

Track whether the action items actually get done. If the same issues keep popping up in future post-mortems, you’re not fixing anything. Success means fewer recurring problems over time.

Q: What’s the one thing most teams skip?

The follow-up checkpoint on Day 14. It’s easy to forget. But that 15-minute stand-up keeps things moving. Without it, action items pile up and nothing gets resolved.

TL;DR: Run a focused, neutral post-mortem within 48 hours of project close. Assign a moderator, stick to a 30-minute agenda, document Lessons Learned, and assign owners to each action. Publish the report by day 3 and schedule a 15-minute follow-up in two weeks to confirm fixes.

What’s happening

A post-mortem review is a structured meeting held shortly after a project ends to inspect what succeeded, what failed, and—most important—how to apply those lessons to the next initiative.

It’s not about pointing fingers. It’s about evidence-based improvement. PMI says these reviews are mandatory for any project lasting more than two weeks or involving three or more teams (as of their 2024 guidelines).

How do you actually run one?

Start by setting the stage on Day 0.

Set the stage (Day 0)
- Send a 3-question survey to every participant: “What went well?”, “What didn’t?”, and “One thing we must change next time.” Stick to Google Forms or Microsoft Forms—just know the 2026 default response limit is set to 100 answers.
- Pick a moderator who wasn’t a core contributor. Neutrality matters here. Then schedule a 30-minute Teams or Zoom call within 48 hours of project closure.

Build the agenda (Day 0, 15 minutes)

Time	Topic	Owner
0:00–0:05	Recap scope & timeline	PM
0:05–0:15	Review survey themes	Moderator
0:15–0:25	Bucket actions: “Keep,” “Stop,” “Try”	Entire group
0:25–0:30	Assign owners & due dates	Moderator

Document & publish (Day 1–3)
- Use the PostHog Post-Mortem Template—it keeps everything consistent.
- Include: project name, timeline, key metrics, survey quotes, action items with owners and deadlines.
- Email the PDF to all stakeholders and archive it in the company Notion workspace under “Retrospectives.”
Follow-up checkpoint (Day 14)
- Schedule a 15-minute Zoom stand-up, but only invite the action owners.
- Update the document with status: “Done,” “In Progress,” or “Blocked.”
- If something’s blocked, escalate to the VP of Engineering within 24 hours.

What if the standard approach falls flat?

Try these alternatives to shake things up.

Anonymous retro: Use Retrium’s silent brainstorming mode to cut down on groupthink. Great for teams where people don’t feel safe speaking up.
Written-only retro: Skip the live meeting entirely. Run a 48-hour Slack thread instead, then post the final report without a call. Handy for teams spread across time zones.
Third-party facilitator: Bring in an external coach from Liberating Structures if internal tensions are making things awkward.

How do you keep this from becoming a one-time thing?

Make retrospectives a habit, not an afterthought.

Run retrospectives every sprint—every two weeks—rather than waiting for a project to wrap up. Scrum Alliance has recommended this cadence since 2025.
Train every new hire on the “Keep / Stop / Try” framework during onboarding. It’s simple but effective.
Set up a shared “Blame-Free Learning Log” in Confluence. It auto-archives every retro, so patterns are easy to spot later.

Who should attend the post-mortem?

Invite everyone who had a hand in the project—plus a few extras.

Core contributors? Absolutely. Stakeholders who can provide context? Yes. Even that one person who joined late but still contributed? Bring them too. The more perspectives, the better.

What’s the biggest mistake teams make?

Turning it into a blame session.

Honestly, this kills psychological safety faster than anything. Keep the focus on what happened, not who messed up. The goal is learning, not finger-pointing.

How detailed should the report get?

Detailed enough to be useful, but not so dense it collects digital dust.

Include the key metrics, survey quotes, and action items with clear owners and deadlines. Skip the novel-length breakdowns—no one’s going to read those.

What if the project was a total disaster?

Run the post-mortem anyway—but adjust your approach.

Maybe swap in anonymous feedback first. Or bring in that third-party facilitator we talked about earlier. The goal isn’t to relive the trauma; it’s to figure out what to do differently next time.

How do you handle pushback from leadership?

Frame it as a risk-reduction exercise, not an extra meeting.

Say something like, “This isn’t busywork. It’s how we avoid repeating the same mistakes.” Most leaders get that.

What tools do you really need?

A survey tool, a template, and a way to share the final report.

Google Forms or Microsoft Forms for the survey. The PostHog template keeps things consistent. Notion or Confluence for archiving. That’s it.

Can you do this remotely?

Absolutely—just tweak the format a bit.

Use breakout rooms for the “Keep / Stop / Try” bucketing. Try a Slack thread for the written-only retro. The key is keeping it interactive, even from a distance.

What if no one shows up?

Make attendance mandatory—or at least make it worth their time.

Schedule it right after the project ends, when everyone’s still in the loop. Or offer to buy them coffee (virtual or real). People are more likely to come when they see the value.

How do you measure success?

Track whether the action items actually get done.

If the same issues keep popping up in future post-mortems, you’re not fixing anything. Success means fewer recurring problems over time.

What’s the one thing most teams skip?

The follow-up checkpoint on Day 14.

It’s easy to forget. But that 15-minute stand-up keeps things moving. Without it, action items pile up and nothing gets resolved.