Every retrospective format in AgileMate, with its steps and timings. Free to read, free to run.
A visual metaphor retro: the wind pushes the team forward, anchors hold it back, rocks are risks ahead, and the island is the goal. Great for teams that think better in pictures.
Four lists — Liked, Learned, Lacked, Longed for — give a balanced view of the sprint. Simple to explain and works for almost any team.
Five arms — Keep, More of, Less of, Start, Stop — push the team beyond good/bad thinking into dosage: what deserves more, what deserves less.
An emotions-first retro: what made us mad, sad, or glad. Great for new teams learning to talk about feelings safely.
The classic action-oriented retro: what should we start doing, stop doing, and continue doing. Fast, concrete, hard to get wrong.
No agenda until the meeting starts: the team proposes topics, votes, and discusses them in strict timeboxes, thumb-voting to continue or move on. Democratic and energizing.
Keep, Add, Less, More — a calm dosage conversation about existing practices rather than dramatic change. Good for stable teams fine-tuning.
Drop, Add, Keep, Improve — a decisive retro that forces clear choices about practices: what dies, what's born, what stays, what evolves.
The minimal retro: what worked (+) and what we'd change (Δ). Two columns, thirty minutes, zero setup — ideal when time is short.
The team is a race car: the engine drives us, the parachute drags us, the abyss ahead is the risk, the bridge is how we cross it.
Hot air lifts us, sandbags weigh us down, the storm ahead threatens, sunny skies are where we want to fly. Adds a forward-looking twist to the classic metaphor retro.
House of straw (fragile), house of sticks (okay-ish), house of bricks (solid) — plus the wolf: what could blow us over. A playful way to audit practices.
The team as a superhero: superpowers, kryptonite, arch-nemesis, and trusty sidekick. Reveals identity and dependencies with humor — best with teams that know each other.
The genie grants three wishes: one for the product, one for the process, one for the team. Wishes reveal what people truly miss — then get grounded into actions.
Review the sprint as a film: title and genre, star rating, best scene, scene that should have been cut, and what the sequel needs. Light-hearted but surprisingly revealing.
Everyone marks the sprint's weather — sunny, cloudy, rainy, stormy — then forecasts the next one. A gentle, low-pressure format for teams new to retros.
Pick the sprint's most painful problem and ask 'why' five times until the root cause appears. Narrow but deep — best when one issue clearly dominates.
An Ishikawa diagram for one stubborn problem: causes sorted along bones like people, process, tools, environment. Structured root-cause hunting for recurring issues.
Three rings — we control, we influence, the soup we swim in. Sorts frustrations by agency and turns each ring into a different kind of response. Antidote to helplessness.
Imagine the next sprint has failed spectacularly — now explain why. Prospective hindsight surfaces risks people hesitate to name directly. Run it before risky sprints.
The whole team wears one thinking mode at a time: facts, feelings, risks, benefits, ideas, and process. Prevents debate ping-pong and gives quiet voices structure.
Reconstruct the sprint on a long wall: events, decisions, surprises, and how people felt at each point. The richest picture of what actually happened — worth the longer timebox.
Everyone draws their personal energy line across the sprint, then the lines are overlaid. Shared dips and peaks point straight at what needs discussing.
Everyone marks their charge level, then the team collects what charges and what drains the battery. A fast energy check that ends with one recharge action.
Each person sums up the sprint in exactly one word, then the words are explored. The fastest retro there is — perfect as a check-in or when time collapsed.
Anonymous vote: is everyone here an Explorer, Shopper, Vacationer, or Prisoner? Measures engagement with the retro itself and opens an honest conversation about it.
Roses bloomed (wins), buds are promising (opportunities), thorns hurt (problems). The bud column makes this gentler and more forward-looking than most three-column retros.
Wishes, Risks, Appreciations, Puzzles — a balanced mix of hopes, dangers, gratitude, and open questions. Appreciations make it a warm mid-project pulse check.
Jump ahead: the release shipped and it was a triumph. Looking back from that future, what did we do to get there? Back-casting makes bold plans feel concrete.
The team picks 5–6 dimensions it cares about — like quality, focus, fun, courage — rates each, and plots a radar. Gaps between ratings spark the honest conversations.
A retro dedicated to gratitude: everyone writes and hands appreciation cards to teammates. Run it after hard sprints or releases — trust is a deliverable too.
The team fills the board over a day, then meets briefly to decide. For teams that are never all awake at once — the call is spent on decisions, not on collecting input.
Twenty minutes on how connected the team felt, not on what it shipped. Distance is not only kilometres, and a distributed team loses the signal long before anyone mentions it.
Trace the last few pieces of work end to end and mark every point where they changed hands. Across time zones a handoff costs a day that never shows up on the board.
A strict rotation so the fastest microphone stops winning. On a call the quiet half of the team disappears unless the format makes room for it.
A distributed team lives inside its tools and pays the rent in minutes. Thirty minutes to find where the tools cost more than they give.
A distributed team quietly splits into smaller ones. This retro finds the seam — a location, a time zone, a channel, a role — and moves it a little.
Walk the board right to left, one card at a time, and let the work tell you where it got stuck. For teams whose problems live between the columns rather than inside them.
Count what each person is really carrying, including the things not on the board. Uncomfortable and short — for teams that finish nothing while everyone is busy.
Sort the in-progress cards by how long they have been sitting there, and talk only about the oldest. Short, blunt, and hard to argue with.
Follow a single item from request to delivered and mark every wait. The most reliable way to show a team that its delays are not where it thinks.
Give every impediment of the last few weeks a name, an owner and a decision. For teams that raise the same blockers in every standup and resolve none of them.
Put two dates on every cross-team dependency — when you found it and when you needed it — and go looking for the signal you already had. For groups that keep learning about the other team's work halfway through their own.
Sit both teams in front of the work that crosses between them and mark the moment it stopped being anybody's. For two teams whose items sit untouched for days after one side considers them delivered.
Put every team's open requests to one shared group on a single wall and look at them with that group in the room. For the platform, data or ops team everyone is waiting on and nobody wants to blame.
Take the cross-team decisions of the last few weeks and go looking for them in the actual work. For groups that leave meetings agreeing with each other and find out months later that nothing changed.
Test the room against the changes it wants before spending an hour on things nobody present can decide. For cross-team retros that produce good conversation and no movement.
Write down what the team believed when it started each piece of work, then check which of those beliefs was ever tested. For teams that ship on time and find out months later that the premise was wrong.
Read out what the team knows about its users and count how much of it anyone present has heard first-hand. For teams whose picture of the user is old enough to be wrong.
Walk through what the team shipped over the last few months and find the parts nobody uses. For teams that only ever add — every unused feature is paid for again in every release.
Put everything the team finished next to what actually changed for the people using it. For teams with a full done column and no way to tell whether any of it mattered.
Find the choice the team made once, long ago, and has been building on ever since without checking. Short and specific — for teams whose constraints are older than their product.
Rebuild what happened as a list of facts everyone agrees on, and refuse to discuss causes until the list is finished. For teams whose incident reviews turn into an argument in the first ten minutes because two people are describing different incidents.
Measure the distance between the moment the system broke and the moment a human knew, then spend the whole session on that number. For teams that repair incidents quickly and still lose a day to them.
Replace the hunt for one root cause with the list of conditions that all had to hold at the same time. For teams whose reviews stop at the last change anyone made and never reach the reasons that change was so easy to get wrong.
Replay the incident from the position of whoever was holding the pager and list everything they had to hunt for, guess, or wake somebody up to ask. For teams that keep hardening the system and never improve the twenty minutes it costs someone at three in the morning.
Collect the times something nearly went wrong and was quietly caught, before one of them stops being caught. For teams with a clean incident record and a quantity of luck they have never counted.