The Went Well, To Improve, Ideas retrospective
A simple, balanced retro: what went well, what to improve, and ideas to try next. How to run it so it ends in experiments, not complaints.
Went Well, To Improve, Ideas has one column to celebrate, one to fix and one to experiment. It's a close relative of the familiar "what went well, what didn't" retro, with one useful change: suggestions get their own column instead of being mixed in with problems.
That matters. In a two-column retro, problems take over and ideas get lost among them. Here, people can write a problem in one column and a possible fix in another, and the discussion has somewhere to go.
The three columns
- Went well: What worked and is worth repeating? Be specific: "The new test environment meant nobody waited for a deploy," not "Good sprint." Specific wins can be repeated on purpose.
- To improve: What fell short or got in our way? Say what happened and what it cost: "Three stories came back from review for missing acceptance criteria." Facts make a better discussion than complaints.
- Ideas: What should we try next? Small experiments: "Write acceptance criteria together in refinement for two sprints." The best ideas often answer a card in To improve.
When to use it
It's a good default for regular sprint retros, teams new to retros, and mixed groups that include people from outside engineering, because the columns need no explanation.
If the team needs decisions about its habits, Start · Stop · Continue frames every card as an action. For more nuance than "went well" and "didn't," try the 4 Ls. Rose · Bud · Thorn has the same shape, with a bud for early signs of promise.
How to run it
- Set the stage (5 minutes). Name the period and review last retro's action items. Did last time's ideas get tried?
- Write alone (7 minutes). Everyone writes in all three columns. Encourage an idea for each improvement, but don't require it.
- Group (5 to 10 minutes). Merge duplicates, then put each idea next to the improvement card it answers.
- Vote (3 minutes). Spend votes on the topics that matter most. Hidden tallies keep people from following the crowd.
- Discuss (20 minutes). Take the topics in order of votes. For a To improve topic, agree on what happened before talking about fixes, then turn to the Ideas cards.
- Agree actions (5 to 10 minutes). Pick one or two ideas to try as experiments, each with an owner, a due date and a way to tell whether it worked.
- Close (3 minutes). Read a few Went well cards aloud. They're easy to forget in a busy sprint.
Tips
- Treat ideas as experiments. "Try it for two sprints, then decide" is easier to agree to than a permanent change, and it gives the next retro something to review.
- Keep To improve factual. "Reviews are slow" becomes useful when it says how slow, and what waited.
- Notice an empty Ideas column. It can mean people don't believe change is possible. Following through on action items is the best answer to that.
Run it in Scrums
Choose New session → Retrospective on your team's page and pick Went Well · To Improve · Ideas from the formats. Everyone writes privately, and other people's cards stay face-down until the facilitator reveals them, so nobody's first card sets the tone. Then group similar cards (or let Auto-group cluster them), vote with hidden tallies, and discuss the topics in order, adding action items with an owner and a due date. Open action items come back in the next retro's Review step, so you can see which ideas were tried. Retros are anonymous by default, and Scrums is free.
Try a retro in your browser, or sign up free.
Questions
How is this different from a "what went well, what didn't" retro?
The two-column version mixes problems and suggestions, and the problems usually win the time. A separate Ideas column keeps suggestions visible, and lets people propose something new that isn't a fix for any problem.
Is it a good format for a first retro?
Yes. The columns need no explanation, the format balances positive and negative, and it fits easily in 45 minutes.
What's the difference between Ideas and action items?
An idea is a suggestion anyone can write. An action item is what the team agrees to do, with an owner and a due date. Most ideas won't become actions, and that's fine: the discussion decides which are worth trying.