Writing Effective Start Stop Continue Feedback
This skill teaches you how to craft clear, specific, and constructive feedback items across the Start, Stop, and Continue categories—making each one behavior-focused, blame-free, and immediately actionable.
To write effective start stop continue feedback, focus each item on a specific, observable behavior rather than a personality trait. Use concrete language that describes what happened, why it matters, and what action to take. Avoid blame or vague suggestions. Each feedback item should be actionable enough that the recipient knows exactly what to do differently starting tomorrow.
Outcome: You produce feedback items that recipients understand, accept, and can act on immediately—leading to measurable behavior change rather than defensiveness or confusion.
Prerequisites
- Basic understanding of the Start Stop Continue framework
- Familiarity with giving and receiving feedback in team or 1-on-1 settings
Overview
The Start Stop Continue framework gives teams a clean structure for organizing feedback, but structure alone doesn't guarantee quality. The real power of the method comes from how each individual feedback item is written. Vague, judgmental, or unfocused items get ignored. Specific, behavior-focused, and actionable items drive real change.
Writing effective start stop continue feedback is a distinct skill that sits between knowing the framework and facilitating a full retrospective. It requires you to translate gut feelings and frustrations into precise, constructive language that a colleague can immediately understand and act on. Whether you're preparing for a team retrospective, a performance review, or an async feedback round, the quality of your written items determines whether the exercise produces transformation or just fills a whiteboard.
This skill covers the principles, patterns, and pitfalls of writing in each of the three categories. You'll learn how to stay behavior-focused, avoid blame traps, calibrate specificity, and test each item for actionability before you share it.
How It Works
Effective start stop continue feedback works because it constrains feedback along two dimensions: category (start, stop, or continue) and quality (specific, behavioral, actionable). The category forces you to commit to a direction—something new, something to eliminate, or something to preserve. The quality criteria force you to move past vague impressions into observable, concrete territory.
Psychologically, behavior-focused feedback is easier to receive than identity-focused feedback. Saying "We should stop scheduling meetings without agendas" is about a practice, not a person. It doesn't trigger defensiveness the way "You're disorganized" does. This distinction—between behavior and identity—is the single most important principle in writing feedback that actually gets implemented.
Actionability is the second critical lever. Each item should pass the "Monday morning test": could someone read this item on Monday morning and know exactly what to do differently that week? If the answer is no, the item needs to be rewritten with more specificity. This test applies equally across all three categories—start items need to describe a concrete new behavior, stop items need to pinpoint the exact practice to eliminate, and continue items need to name the specific behavior worth preserving so it doesn't accidentally get dropped.
Step-by-Step
Step 1: Gather Raw Observations Before Writing
Before you start crafting feedback items, spend 5-10 minutes jotting down raw observations from the period you're reflecting on. These can be messy—frustrations, highlights, things that felt off, moments that went well. Don't censor yourself yet. The goal is to create a pool of raw material you'll refine in later steps.
Review meeting notes, project outcomes, Slack threads, or your own journal entries. Look for patterns: recurring friction, repeated wins, or new experiments that worked. Having concrete reference points prevents you from writing feedback based on recency bias or a single emotional moment.
Tip: Set a timer for 5 minutes and write continuously without editing. You'll sort and refine later—quantity of raw observations matters more than polish at this stage.
Step 2: Sort Observations into Start, Stop, and Continue Buckets
Take your raw observations and assign each to a category. Ask yourself three questions for each observation:
- Start: Is there a new behavior, practice, or habit that would address this observation?
- Stop: Is there an existing behavior or practice causing this problem that should be eliminated?
- Continue: Is there something already working well that this observation highlights?
Some observations might fit multiple categories. A frustration about unclear requirements might generate both a "stop" item (stop starting sprints without acceptance criteria) and a "start" item (start writing user stories with defined done criteria). That's fine—separate them into distinct items.
Tip: If you're struggling to categorize an observation, it's often a sign the observation is too abstract. Try making it more specific—name the meeting, the process, or the interaction—and the right category usually becomes obvious.
Step 3: Rewrite Each Item Using Behavior-Focused Language
Now refine each item so it describes an observable behavior rather than a personality trait, attitude, or vague concept. Use the formula: [Action verb] + [specific behavior] + [context or frequency].
For example, transform "Be more communicative" into "Start posting daily async standup updates in the #project channel by 10 AM." Transform "Stop being negative" into "Stop raising objections during brainstorming sessions—save critical evaluation for the analysis phase."
Behavior-focused language answers the question "What would a camera see?" If a video camera couldn't record the behavior you're describing, it's too abstract. "Being more proactive" is invisible to a camera. "Flagging blockers within 24 hours of identifying them" is observable and recordable.
Tip: Read each item aloud and ask: "Could two different people interpret this the same way?" If not, add specificity until the interpretation converges.
Step 4: Remove Blame and Judgment from Each Item
Review every item for hidden blame, sarcasm, or judgment. Blame often sneaks in through adverbs ("finally start..."), absolutes ("always," "never"), and loaded adjectives ("lazy approach," "careless mistakes"). These words make the recipient defensive and undermine the actionability of your feedback.
Replace blame with impact. Instead of "Stop wasting everyone's time with unnecessary meetings," write "Stop scheduling recurring meetings that don't have a clear agenda or decision to make—this would free up approximately 3 hours per week for the team." The second version communicates the same concern but frames it around impact rather than accusation.
For "continue" items, be careful not to backhanded-compliment: "Continue actually showing up on time" implies past failure. Simply write "Continue arriving on time to standup—it keeps us on schedule."
Tip: A simple test: would you feel comfortable if the item were projected on a screen with your name attached during an all-hands meeting? If not, revise the tone.
Step 5: Apply the Monday Morning Test for Actionability
Read each item and ask: "If someone received this on Friday afternoon, would they know exactly what to do differently on Monday morning?" This is the actionability litmus test.
Items that fail this test are usually too broad. "Start improving our documentation" fails—it doesn't specify what documentation, what improvement looks like, or where to begin. "Start adding a README with setup instructions to every new repository before requesting code review" passes—it names the artifact, the content, and the trigger.
For "stop" items, actionability means being specific about what replaces the stopped behavior. "Stop using email for urgent requests" is better as "Stop using email for urgent requests—use the #urgent Slack channel instead so the on-call person sees it within 15 minutes."
Tip: If an item feels too large for one action, break it into smaller items. Two specific items are always more useful than one ambitious but vague one.
Step 6: Balance Your Feedback Across All Three Categories
Review your complete set of feedback items. A common trap is writing mostly "stop" items (a complaint list) or mostly "start" items (a wish list) while neglecting "continue" items. Aim for balance—not necessarily equal numbers, but genuine representation of all three categories.
"Continue" items are especially important because they reinforce positive behaviors that might otherwise be dropped during change. Teams that only hear about what to fix lose sight of what's already working. Make your continue items just as specific and behavior-focused as your start and stop items—"Continue doing good work" is worthless, while "Continue pairing on complex pull requests before merging—it's caught three significant bugs this sprint" is powerful.
If you genuinely can't think of continue items, you may be writing from a place of frustration rather than balanced reflection. Step back and look for bright spots before finalizing.
Tip: A ratio of roughly 2 start/stop items to every 1 continue item works well for most teams. Adjust based on context—a team in crisis may need more stop items, while a high-performing team may need more continue items.
Step 7: Review and Finalize Before Sharing
Before submitting or sharing your feedback, do a final pass with these checks:
- Specificity check: Does each item name a concrete behavior, practice, or process?
- Blame check: Is each item free of accusatory language, sarcasm, and hidden judgment?
- Actionability check: Does each item pass the Monday morning test?
- Ownership check: Is it clear who should act on each item (a specific role, the whole team, or a named individual if appropriate)?
- Overlap check: Are any items redundant? Merge or eliminate duplicates.
If you're writing for an anonymous retrospective, ensure your items don't accidentally reveal someone else's identity through overly specific scenarios. If you're writing for a 1-on-1 or performance review, consider adding a brief explanation of why each item matters to you—this context helps the recipient prioritize.
Examples
Example: Engineering Team Sprint Retrospective
A software engineering team of six is wrapping up a two-week sprint. The sprint had a rocky deployment, some communication gaps, and one practice that went particularly well. A team member is preparing their start stop continue feedback items for the retro.
Raw observations: Deployment failed on Wednesday because nobody checked staging. Daily standups ran 30+ minutes several times. The new practice of pairing on complex PRs caught two critical bugs.
Refined feedback items:
- Start: Start running a smoke test on staging and posting results in #deploys at least 2 hours before any production deployment. This gives us a safety net and time to fix issues before they hit production.
- Stop: Stop allowing standup discussions to go beyond 2 minutes per person. If a topic needs more time, post it in a parking lot and schedule a 15-minute follow-up with only the relevant people.
- Continue: Continue pairing on pull requests that touch more than 3 files or modify shared utilities. This sprint, pairing caught two bugs that would have reached production—one of which affected billing calculations.
Each item names a specific behavior, includes context for why it matters, and is actionable enough that the team could implement it starting next sprint.
Example: Manager Writing Feedback for a Direct Report's 1-on-1
A marketing manager is preparing start stop continue feedback for a quarterly 1-on-1 with a content strategist. The strategist produces great work but struggles with deadline communication and has recently adopted a useful new practice.
Refined feedback items:
- Start: Start sending a brief status update in Slack every Friday by 4 PM listing which deliverables are on track and which might slip. This helps me adjust priorities before the weekend and avoid Monday surprises.
- Stop: Stop agreeing to deadline commitments in meetings without checking your current workload first. Instead, say 'Let me check my calendar and confirm by end of day.' This will reduce the number of missed deadlines, which have averaged two per month this quarter.
- Continue: Continue including competitive analysis sections in your content briefs. The sales team has specifically called out these sections as useful in three different deal reviews this quarter—they're using your research in prospect conversations.
Notice that each item is addressed to the individual, framed around behavior (not character), includes supporting evidence, and passes the Monday morning test. The stop item even includes replacement language the strategist can use.
Example: Transforming Weak Feedback into Strong Feedback
A team member has drafted start stop continue feedback items but they're vague and blame-laden. This example shows the before-and-after of applying the writing principles.
Before (weak):
- Start: Be more organized.
- Stop: Wasting time in meetings.
- Continue: Keep up the good work.
After (strong):
- Start: Start creating a shared agenda document at least 24 hours before each team meeting, and pin it in the meeting's Slack channel so attendees can add topics in advance.
- Stop: Stop scheduling 60-minute meetings as the default. Use 25 minutes for discussions and 50 minutes for workshops—this forces tighter facilitation and frees up buffer time between calls.
- Continue: Continue sharing customer interview recordings in the #research channel within 24 hours of each call. Three team members have told me they watch every one, and it's visibly improving how we talk about user needs in planning.
The weak versions are vague, judgmental, and unactionable. The strong versions name specific behaviors, include context, and tell the reader exactly what to do.
Best Practices
Use one feedback item per behavior—don't bundle multiple observations into a single bullet point, as compound items are harder to act on and easier to partially ignore.
Ground each item in a specific, recent example whenever possible. "Start documenting decisions in our Confluence space after each architecture review" is stronger than "Start documenting more."
Write continue items with the same specificity as start and stop items—vague praise doesn't reinforce behavior as effectively as naming exactly what someone did well.
Frame stop items in terms of impact on outcomes ("this delays our release cycle by 2 days") rather than impact on your feelings, which keeps the conversation professional and data-oriented.
When writing start items, suggest a concrete first step or trigger point ("at the beginning of each sprint planning" or "before merging any PR") so the new behavior has a natural place in existing workflows.
Timebox your writing to 15-20 minutes per session. Feedback written during extended sessions often becomes over-edited or loses its authentic voice.
Common Mistakes
Writing feedback about personality traits instead of behaviors
Correction
Replace trait-based language ('be more proactive,' 'stop being negative') with observable actions ('start flagging blockers in standup within 24 hours,' 'stop raising objections during the brainstorm phase—save them for the evaluation round'). If a camera can't record it, rewrite it.
Making continue items generic or throwaway
Correction
Treat continue items as seriously as start and stop items. Instead of 'continue the good work,' write 'continue running 15-minute post-incident reviews within 48 hours of each outage—these have cut our repeat incident rate by 40%.' Specific continue items prevent valuable practices from being accidentally dropped.
Writing items that are actually complaints disguised as feedback
Correction
Test each item by asking: 'Does this tell someone what to do, or just what I'm unhappy about?' Transform complaints ('Our standups are too long') into actionable items ('Stop allowing standup discussions to exceed 15 minutes—move detailed conversations to a follow-up thread').
Bundling multiple issues into a single feedback item
Correction
Split compound items into separate, focused pieces. 'Start writing tests and also improve our deployment process and update the wiki' is three separate items. Each one deserves its own line so it can be discussed, prioritized, and tracked independently.
Using start stop continue feedback as an anonymous way to target individuals
Correction
Keep team-level feedback about team-level behaviors and processes. If feedback is genuinely about one person's performance, deliver it in a private 1-on-1 conversation, not through a team retrospective board where it becomes a public callout.
Other Skills in This Method
Facilitating Start Stop Continue Retrospectives
How to plan, run, and timebox an effective Start Stop Continue retrospective session with your agile team, from setting ground rules to closing with action items.
Using Start Stop Continue in Performance Reviews and 1-on-1 Meetings
How to adapt the Start Stop Continue framework for individual performance conversations, manager check-ins, and self-reflection outside of team retrospectives.
Running Start Stop Continue as a Team Icebreaker Activity
How to use a lightweight Start Stop Continue exercise as a warm-up or icebreaker to build psychological safety and normalize giving feedback in new or forming teams.
Categorizing and Prioritizing Feedback Items
How to sort, group, and dot-vote on collected Start, Stop, and Continue items to identify the highest-impact actions the team should commit to next.
Building Start Stop Continue Templates and Worksheets
How to design reusable templates, worksheets, and digital boards (Miro, Google Docs, Notion) that structure the feedback collection and make sessions efficient.
Crafting Actionable Start Stop Continue Questions and Prompts
How to write and select targeted questions for each Start, Stop, and Continue category that elicit specific, constructive, and actionable feedback rather than vague responses.
Frequently Asked Questions
How many start stop continue feedback items should I write?
Aim for 2-4 items per category for a team retrospective, or 1-2 per category for a 1-on-1. Quality matters more than quantity—three well-written, specific items drive more change than ten vague ones. If you have many items, prioritize the ones with the highest potential impact.
How do I write start stop continue feedback without blaming someone?
Focus on the behavior or process, not the person. Use phrases like 'stop scheduling meetings without agendas' rather than 'stop being unprepared.' Describe impact on outcomes ('this delays releases') rather than intent ('you don't care about deadlines'). If you read your item and it sounds like an accusation, rewrite it around the observable action.
What's the difference between start stop continue feedback and pros and cons lists?
Pros and cons describe a current state—what's good and what's bad. Start stop continue feedback is action-oriented: it tells people what to do next. Each item should include a clear direction (begin, cease, or maintain) and enough specificity to act on, making it a planning tool rather than just an evaluation.
Can I use start stop continue feedback for self-reflection?
Absolutely. Apply the same principles—write behavior-focused, specific items directed at yourself. For example, 'Start blocking 30 minutes each morning for deep work before checking email' is more useful than 'Be more focused.' The framework works well in personal journals, quarterly self-reviews, and goal-setting sessions.
How do I handle start stop continue feedback that's too vague to act on?
Ask clarifying questions: 'Can you give me a recent example?' or 'What would this look like in practice?' If you're reviewing your own items before submitting, apply the Monday morning test—if someone couldn't act on it first thing next week, add specificity about the behavior, the context, and the trigger.
Should start stop continue feedback be anonymous?
It depends on your team's psychological safety. Anonymous feedback lowers the barrier to honesty but makes follow-up conversations harder. If your team is new to the practice, anonymous collection followed by group discussion is a solid approach. As trust builds, named feedback enables richer dialogue. Either way, the writing quality principles remain the same.