I've been in enough retrospectives to notice a pattern: after a while, they start feeling repetitive.
The retros aren't pointless. Teams just lose, bit by bit, the things that made them useful in the first place.

You get the same discussions, the same friction points, and every so often the same action items as the retro before.
People haven't stopped caring. The process has eroded underneath them.

Retrospective.fun grew out of wanting to fix the things I kept running into as a facilitator and a participant.

Below is the thinking behind the product, and the problems I wanted it to solve.

The real reasons retros go stale

1. Retros have bad memory

Most tools (and many teams) treat retros as isolated events.
Notes disappear into screenshots or a tool nobody revisits.
The team has conversations, but the system forgets everything.

So you rediscover the same problems every few sprints.

2. Facilitators carry too much overhead

Grouping notes, wrangling votes, keeping the timer alive, managing the order of discussion topics. It adds up.
If the tool doesn't take some of that load, the facilitator ends up split between logistics and actually listening to the team.

3. There's no link between "talked about" and "did something about it"

This one is subtler.
You can create action items, but they rarely feel tied to the discussion that produced them.
And come the next retro, there's no sense of what keeps recurring. The context is gone.

Those three are what I built around.

A different approach: structure first, then automation

Retrospective.fun uses a template-driven, stage-based workflow.
A retro works best when everyone knows what phase they're in and what that phase is for.

Templates define:

  • what the team should be doing
  • who can do what
  • when you can move forward
  • how the flow is structured

It's predictable in a good way. You don't lose momentum halfway through, because the process carries you.

For facilitators, this simplifies the job.
For participants, it reduces confusion.

Grouping and prioritisation that doesn't feel like work

I didn't want a tool where grouping notes feels like doing chores.

So I built two things:

1. Grouping that works with or without AI

You can group manually with drag and drop, or, if you're on the AI tier, let the tool propose clusters and refine from there.

It's fast, and you keep full control.

2. A proper discussion queue

Votes lead straight into a queue of topics to work through.
You always know:

  • what's next
  • how many items are left
  • where the team disagrees
  • when you're done

It sounds basic. In practice it removes a lot of friction.

Action items that actually connect back to the session

Action items keep their context:

  • which discussion created them
  • which notes they're tied to
  • which session they came from
  • whether this is a recurring theme

Over time you get a sense of patterns instead of a pile of isolated tasks.
That mattered to me because I wanted a tool that didn't throw away the team's collective memory.

AI that stays in its lane

The AI features aren't there to run the retro. They're there to remove repetitive steps:

  • suggesting groupings
  • summarising what was discussed
  • highlighting patterns across past sessions

There's a reliability layer behind all of it (caching, rate limits, timeouts) so the tool doesn't spam the API or run off unsupervised.
The part that matters more: the AI never decides anything. The team does.

Built for smaller, real teams

Retrospective.fun is built for the kind of teams I've worked in: small and mid-sized, practical, busy, and not interested in ceremony. It isn't aimed at large-scale "transformation programs".

The pricing reflects that:

  • Free for small teams
  • Pro for better retention and custom templates
  • Pro+ if you want all the AI and insights

No pressure and no large commitments.

What changes when retros stop being admin-heavy

When the tool carries the overhead and the memory, the retro starts feeling like it did when it actually worked. Real issues surface earlier, the team stays on them, and you stop rediscovering last quarter's problems.

That's all I was after: less friction, and a team that gets something back for the hour it spends together.