Winning - Puzzling - Improving
Works well for larger groups: highlight wins, open questions, and improvements.
Winning
Outcomes and behaviors that felt like clear wins.
Puzzling
Things you do not fully understand yet—need discovery.
Improving
Areas you already know need better execution or support.
This board is for demo purposes only. Your responses are not saved. Close or refresh the page to clear all cards. Do not add any sensitive information.
What is Winning - Puzzling - Improving?
Winning - Puzzling - Improving is a three-column retrospective format designed to balance celebration, curiosity, and action. The format was developed by agile coaches working with larger teams who observed that standard retro formats often skip the exploration phase—jumping from problem identification to solution without understanding the problem fully. The "Puzzling" column addresses this gap by creating explicit space for things the team does not yet understand.
The three columns represent three states of knowledge. "Winning" captures things the team knows are working—clear victories with understood causes. "Puzzling" holds things the team does not fully understand—anomalies, surprising outcomes, unclear patterns, or questions without answers. "Improving" identifies areas where the team knows what needs to change and can articulate the direction, even if the specific solution is not yet defined.
The "Puzzling" column is what makes this format distinctive. Most retrospective formats assume the team can clearly categorize everything as good or bad. In reality, many sprint outcomes are genuinely confusing: a feature that shipped on time but received no user adoption, a process that felt painful but produced great results, or a team dynamic that suddenly shifted without apparent cause. Puzzles deserve investigation before improvement.
When to use Winning - Puzzling - Improving
This format excels with larger teams of eight to fifteen people where the diversity of perspectives means that what is puzzling to one person may be understood by another. The cross-pollination of insights that happens when someone explains a puzzle and another person says "oh, I know why that happened" is one of the most valuable moments in any retrospective. Larger groups increase the probability of these breakthrough connections.
Use this format when the team is facing ambiguity—new markets, unfamiliar technology, organizational changes, or any situation where the team needs to make sense of their experience before improving it. It is particularly valuable at the end of a spike or exploration sprint where much was learned but not all lessons are clear.
The format fits well into a 45 to 60 minute session and works as a regular sprint retro format, especially for teams that value learning and sense-making alongside practical improvement. Avoid it when the team needs to focus exclusively on action items—the Puzzling column can consume significant time in exploration mode that some teams may perceive as unproductive if they are under acute delivery pressure.
How to facilitate Winning - Puzzling - Improving
Start with a brief share: "What is one thing from this sprint that surprised you?" This opening question primes the team for the curiosity mindset that the Puzzling column requires. Then give six to eight minutes for silent writing across all three columns. Encourage participants to be honest about what they genuinely do not understand—there is no expectation that everyone should have all the answers.
Process Winning first with brief celebrations. Then spend the most time on Puzzling. For each puzzle, ask the group: "Does anyone have a theory about why this happened?" Often, the person who experienced the puzzle and the person who can explain it are different people, and the retro becomes the venue for this knowledge transfer. For puzzles that remain unsolved, decide whether they need investigation (assign someone to dig deeper) or acceptance (acknowledge the uncertainty and move on).
Finish with Improving, which should be relatively straightforward if the Puzzling discussion has resolved some ambiguities. Convert the top improvement themes into specific action items. Close by asking: "Did any puzzles get solved today? Did any new puzzles emerge?" This closing question reinforces the learning orientation that makes this format valuable.
Tips for getting the most out of Winning - Puzzling - Improving
Protect the Puzzling column from premature resolution. When someone shares a puzzle, others often jump to explanations or solutions. As a facilitator, slow this down: "Before we solve it, let us make sure we understand the puzzle fully. Can you describe what you expected to happen versus what actually happened?" This discipline ensures the team is solving the right puzzle rather than a simplified version of it.
The Winning column serves a strategic purpose beyond celebration. Wins that the team can explain create a playbook for future success. When someone shares a win, always ask: "Why did this work? Can we replicate the conditions?" Unexplained wins are actually puzzles in disguise—if you do not know why something worked, you cannot deliberately recreate it. Move unexplained wins to the Puzzling column for deeper exploration.
For larger groups, use a "gallery walk" format for the Puzzling column. Post all puzzles on a wall or virtual board and have people silently walk around, adding theories or related observations as sticky notes. This parallel processing approach is more efficient than sequential discussion and generates richer connections because people can see and build on each other ideas in real time.
Variations and adaptations
For remote teams, the Puzzling column benefits from pre-session preparation. Ask team members to submit their puzzles 24 hours before the retro so the facilitator can cluster related puzzles and invite specific people who might have relevant context. This preparation makes the synchronous discussion more efficient because the right people are ready to contribute to each puzzle.
For async teams, turn Puzzling into an ongoing investigation board. Keep it open throughout the sprint so that team members can post puzzles as they encounter them and others can add theories or evidence at their own pace. By the end of the sprint, some puzzles will be resolved through this collaborative investigation, and the synchronous retro can focus on the remaining mysteries.
A powerful variation for data-driven teams adds a "Measuring" column—things the team wants to quantify before deciding whether they are winning or need improving. This bridges the gap between subjective puzzlement and objective improvement by inserting a measurement step. Another adaptation for cross-team retrospectives uses Puzzling as the primary column, collecting puzzles about how teams interact, share knowledge, and coordinate. The cross-team puzzle-solving session often reveals structural issues that are invisible within a single team.
Run Retrospectives in CodeKudu
CodeKudu includes dozens of retrospective board templates, anonymous feedback, AI summaries, and action items that sync to GitHub Issues, Jira, and Linear.
