Start, Stop, Continue retrospective template
The fastest way to turn team feedback into concrete behavior changes.
Start, Stop, Continue is a retrospective format where the team answers three questions: what should we start doing, what should we stop doing, and what should we continue doing. It is the most action-oriented retro template because every card is already a behavior change, not just an observation.
Because the three prompts map directly to decisions, this format works especially well for teams that feel their retros produce discussion but no follow-through. It needs no explanation for new participants and runs comfortably in 30–45 minutes.
Timebox
30–45 minutes
Team size
3–12 people
Cadence
Every sprint, or monthly for non-sprint teams
The Start, Stop, Continue board
Start
What should we begin doing that we are not doing today?
- Start doing 15-minute PR reviews within one business day
- Start writing acceptance criteria before the sprint starts
- Start inviting the designer to backlog refinement
- Start recording decisions in a shared decision log
- Start a rotating on-call role so interruptions hit one person, not everyone
- Start estimating spikes so research work is visible on the board
Stop
What is costing us more than it gives us?
- Stop scheduling meetings during the team’s deep-work block
- Stop merging PRs without a second reviewer
- Stop carrying over tickets sprint after sprint without re-scoping them
- Stop discussing solutions in standup — park them for after
- Stop starting new work when the sprint is already at capacity
- Stop treating flaky tests as “just rerun it”
Continue
What is working well that we should protect?
- Continue the Friday demo — stakeholder feedback has been faster
- Continue pairing on gnarly tickets instead of hero-mode solo work
- Continue keeping standups under 10 minutes
- Continue writing post-incident notes while context is fresh
- Continue the no-meeting Wednesday experiment
- Continue celebrating shipped work in the team channel
How to run a Start, Stop, Continue retrospective
- 1
Set the stage (5 min)
Open the board, restate the sprint or period under review, and remind the team the goal is behavior changes, not blame. Review action items from the previous retrospective first.
- 2
Silent writing (8–10 min)
Everyone adds cards to Start, Stop, and Continue privately. Keep cards hidden during writing so the first loud opinion does not anchor the room.
- 3
Reveal and group (5 min)
Reveal all cards at once, then group duplicates and related cards into themes. The facilitator reads groups aloud without debating them yet.
- 4
Vote (3 min)
Each person gets 3–5 votes to spend on the themes that matter most. Use blind voting so votes are revealed only after everyone has voted.
- 5
Discuss top themes (10–15 min)
Work through the top 2–3 voted themes. For each, agree on what specifically changes next sprint and who owns it.
- 6
Close with action items (5 min)
Convert decisions into action items with an owner and a due date. Review them at the start of the next retrospective — this is what makes Start Stop Continue actually change behavior.
Start Stop Continue examples for engineering teams
Start
- Start adding a rollback plan to every deployment ticket
- Start tagging tech-debt tickets so we can see the trend
- Start doing a 30-minute bug triage before sprint planning
- Start writing one-line ADRs for architecture decisions
Stop
- Stop deploying on Friday afternoons
- Stop reviewing 1,000-line PRs — split them
- Stop letting the same person always take the hardest ticket
- Stop skipping the definition of done when the deadline is close
Continue
- Continue the shared test-fixing hour every Tuesday
- Continue writing runbooks for every new service
- Continue asking "do we need this meeting" before booking it
- Continue post-release monitoring shifts
Start Stop Continue examples for product and design teams
Start
- Start sharing user-research clips in the team channel weekly
- Start writing a one-pager before any feature over two weeks
- Start including support tickets in discovery input
- Start usability-testing prototypes before the sprint, not during
Stop
- Stop changing acceptance criteria mid-sprint
- Stop presenting roadmaps as commitments instead of bets
- Stop skipping design review when "it’s just a small change"
- Stop measuring output (features shipped) without outcome (usage)
Continue
- Continue the monthly win/loss review with sales
- Continue pairing PM and engineer on scoping calls
- Continue publishing release notes users can actually read
- Continue the design crit — quality has visibly improved
Start Stop Continue examples for managers and leadership teams
Start
- Start sharing the "why" behind priority changes in writing
- Start skip-level 1:1s once a quarter
- Start protecting at least one no-meeting day per week
- Start closing the loop on employee survey feedback publicly
Stop
- Stop rescheduling 1:1s at the last minute
- Stop assigning work in DMs where the team cannot see it
- Stop celebrating overtime as commitment
- Stop running status meetings that could be a dashboard
Continue
- Continue giving specific, timely praise in public
- Continue quarterly team health checks
- Continue defending the team’s focus time from external asks
- Continue rotating who runs the team meeting
Start Stop Continue examples for personal performance reviews
The format also works for individual feedback — in 1:1s, 360 reviews, or self-reviews — because it separates new habits, harmful habits, and strengths.
Start
- Start asking for help before being blocked for a full day
- Start writing a short weekly summary of what shipped
- Start delegating tasks that others can grow from
Stop
- Stop rewriting teammates’ code instead of leaving review comments
- Stop saying yes to every request without checking capacity
- Stop waiting for the review cycle to share critical feedback
Continue
- Continue mentoring the two junior engineers — it is visibly working
- Continue raising risks early, even when it is uncomfortable
- Continue the clear, well-structured demo presentations
Best for
- Teams whose retros end in vague agreement instead of decisions
- New teams that need a format with zero learning curve
- Process-heavy environments where habits need explicit pruning
- Quarterly reviews of ways of working, not just sprint mechanics
Facilitation tips
Timebox Stop hardest
Stop is where frustration lives, so it attracts the most discussion. Give it energy, but do not let it consume Continue — protecting what works is how teams avoid re-breaking things.
Push cards toward specificity
"Stop having bad meetings" is a complaint. "Stop booking meetings without an agenda" is a decision. As facilitator, ask "what would that look like on Monday?" until cards are concrete.
Limit to 2–3 changes per retro
A team that starts five things and stops five things changes nothing. Vote, pick the top items, and park the rest for next time.
Review last retro’s items first
Nothing kills retro credibility faster than action items nobody looked at again. Open every session by checking the status of the previous items.
Variations
Start, Stop, Continue, More, Less
Adds "more of" and "less of" columns for behaviors that need tuning rather than switching on or off. Useful for mature teams whose problems are about degree, not direction.
Stop, Start, Continue (Stop first)
Running Stop first drains frustration early so Start and Continue get honest attention. Good for teams coming off a rough sprint.
Async Start Stop Continue
Distributed teams add cards asynchronously over 1–2 days, then meet for 20 minutes only to vote and decide. Works well across time zones.
Frequently asked questions
What is a Start Stop Continue retrospective?+
It is a retrospective format where the team collects feedback in three columns: things to start doing, things to stop doing, and things to continue doing. Each card proposes a concrete behavior change, which makes this one of the most action-oriented retro formats.
How long should a Start Stop Continue session take?+
Plan 30–45 minutes for a team of up to 10: about 10 minutes of silent writing, 5 minutes of grouping, 3 minutes of voting, and 15–20 minutes of discussion and action items.
When should I use Start Stop Continue instead of other formats?+
Use it when you want decisions, not exploration. If the team needs to vent or dig into feelings after a hard sprint, a format like Mad Sad Glad or Rose Bud Thorn surfaces emotion better; Start Stop Continue is best when the team is ready to change how it works.
Can Start Stop Continue be used for personal feedback?+
Yes. It works well in 1:1s, self-reviews, and 360 feedback because it frames feedback as specific habits to start, stop, or keep — which feels more actionable and less personal than trait-based criticism.
Run Start, Stop, Continue online in FocusRetro
The Start, Stop, Continue template is built in: pick it when creating a board, invite the team with a link, and run the phases with timers and blind voting. Free plan available at focusretro.com.
Open the template freeNo credit card required
Related guides
Keep, Problem, Try retrospective template
Three columns that chain together: what to keep, what hurts, and the experiment you run next.
Open the Box retrospective template
Treat how your team works like a box of tools — then decide what stays, goes, and gets added.
Sailboat retrospective template
Wind, anchors, rocks, and the island — one picture of what moves the team and what holds it back.
How to run a sprint retrospective (step-by-step guide)
A step-by-step sprint retrospective guide: the 5-phase agenda, exact timings for a 60-minute session, facilitation tactics, remote tips, and common failure modes.
What is a retrospective? The complete guide for teams
A retrospective is a recurring team meeting for inspecting how you work and deciding what to change. Learn the formats, agenda, timings, and mistakes to avoid.
Retrospective timebox calculator
How long Start, Stop, Continue should run for your sprint length and team size, split across the five phases.
Icebreaker and check-in questions
Openers for the first two minutes of a Start, Stop, Continue session, sorted by situation.