GembaKai readout

Recommended owner: Coach, scrum master or tech lead familiar with the waste walk methodology and the team’s sprint history. The facilitator’s job is to read the data with the team — not to advocate for any particular conclusion before the team sees the numbers.

Goals

What the team will do during the session:

1. Walk the GembaKai readout end-to-end with the team.
2. Identify the dominant categories of non-value-add activity using the data.
3. Quantify the productivity opportunity if those categories were halved or removed.
4. Separate what’s inside the team’s control from what needs external influence.
5. Commit to running at least one follow-on workshop targeting the 80%.

Impact

What the team walks away with:

1. A shared, data-grounded view of where sprint effort actually went — not where anyone thinks it went.
2. Vocabulary and numbers for discussing waste outside the team — a defensible business case for change.
3. A named workshop, a date and a facilitator for the next session.
4. Awareness of which improvements the team can act on directly versus those that need a wider conversation.

Audience and personas

Facilitator

Coach, scrum master or tech lead familiar with the readout and the team’s recent activity. Neutral in framing — surfaces the data and lets the team draw the conclusion.

Required participants

• Every team member who logged activity during the sprint (engineers, PO, BA, QA and designers).
• Anyone whose work appears in the timeline view, even part-time contributors.

Optional but valuable

• Line manager or product manager — particularly when the readout is about to support a business case for change outside the team.
• A peer team’s tech lead or coach as a guest observer (their team will run their own readout next sprint).

Roles

Assign these before the session so everyone knows their job.

The time-keeper and the scribe operate across every segment — the agenda’s Leader column credits whoever is driving the discussion, but those two roles are running underneath the whole hour.

Pre-homework

Send the day before. The session is too short to absorb the data cold.

Facilitator (~30 min)

• Open the GembaKai readout for the sprint. Spot-check the readout for anything that looks anomalous — a category that spiked, a comment that surprised you, a stereotype that’s new.

Tip: You can launch the readout for one sprint (as opposed to the organization as a whole) by selecting your sprint and clicking the “View readout” button in the upper right corner.

• Open the sprint timeline view and the activity view. Pull two or three illustrative comments — especially comments that flagged out-of-sprint or non-value-add work — to have ready when the team asks “where did this number come from?”

• Decide which pages deserve dwell time and which are a quick pass. The readout is eleven pages in twenty minutes — that’s under two minutes per page on average, but you’ll want to spend longer on the dominant categories.

Participants (~10 min)

• Open the sprint activity view and skim your own logged activity. Re-read any comments you wrote.
• Bring one observation: a moment in the sprint where you felt your time was being spent on the wrong thing, or a moment where it wasn’t.
• We’ll start the workshop with everyone placing their observation on the whiteboard.

Important: Closed sprints are usually hidden. You can see all sprints (including closed ones) using the “show closed” toggle in the sidebar. See inset.

Materials and setup

• The GembaKai readout — projected or shared. The readout covers sprint context, the value-add vs. non-value-add split, one page each on the four primary categories (rework, automation, planning and context switching), top-line ROI, a comparison to peer teams and a closing recommendations page.

• The sprint admin pages — timeline view for chronological drill-down and activity view for filtering on stereotype and reading comments.
• The GembaKai readout board, a digital, collaborative whiteboard the team will use to categorize activities and prioritize actions. (You can also use a two-column whiteboard or shared doc labelled Inside our control and Inside our influence if you aren’t using the GembaKai platform).

• A short capture template for the workshop selection (see “Capturing results”).

Agenda (60 min)

Workshop exercises

GembaKai provides a collaborative, self-driving workshop board. The board walks the team through the exercise of categorizing and prioritizing

Facilitation with the GembaKai board

The GembaKai board is a collaborative whiteboard the entire team can use throughout this exercise. As the facilitator, you have the ability to “drive” the board. This means actions you take will be mirrored on every participants view of the whiteboard:

• Only team leads have the option to “drive” (to avoid confusion about who’s driving). If there are multiple team leads, you can relinquish control to another team lead.
• As you pilot the team from the first activity (“Walkthrough”) to subsequent activities (“Drill-down,” “Vote,” and so on) their boards will follow along with you.
• As the facilitator, you have some unique actions available to you. One is to start short countdown timers, such as the 2:00 minute “pre-homework” timer shown above. Clicking the double-chevron starts the 2:00 minute timer, and when it runs down the countdown resumes the workshop timer.
• You can also trigger a team vote. Starting a team vote puts everyone into “voting” mode (you end the vote by clicking on the “Stop voting” menu item).

Exercise 1 — Observation and readout walk-through (20 min)

Start by asking everyone to place their observations (from the pre-homework) on the board. Each team member uses the speed dial in the lower right corner of the screen to place a new card on the board. You can use the 2:00 minute countdown timer to facilitate the exercise and get the team actively involved.

Once everyone has placed their observation, move on to “Walk the readout.”

Walk all eleven pages of the readout with the team. The facilitator’s only goal is to surface the data — not to propose a workshop yet.

How to walk each page

1. Read the headline number aloud.
2. Ask the team: “Does that match what you felt during the sprint?”
3. If the team disagrees with the data, capture it as a side note — it’s a signal for the next sprint’s logging discipline, not a reason to dismiss the readout.

What to dwell on, what to skim

Linger on the four primary categories — rework, automation, planning and context switching. Spend the most time on whichever one or two carry the dominant hours; that’s where Exercise 2 is going to land. Context switching is intentionally given the lightest treatment — it’s a real cost but rarely the dominant lever, and it’s easier to address as a multiplier on the other categories than as a target on its own.

Watch for: the team trying to explain or defend numbers (“yeah but that sprint was unusual”). Note the explanation, but keep moving — the readout is a snapshot, not a verdict.

Exercise 2 — Category drill-down (15 min)

Pick the top one or two categories by hours and dig into them with the team. The system will have identified the four most impactful categories and sorted them for you. For each:

1. What does the readout say? Total hours and how it breaks down inside the category (e.g., rework split by item, automation split by stereotype).
2. What do the comments say? Look at the comment cards and drag them onto the board, focusing on what activities caused the most drain on team productivity. Read three or four representative comments aloud — particularly any that flagged “this should have been automated” or “this was rework caused by X.” The team’s own words from the sprint are usually more vivid than the aggregate.
3. What’s the recovery potential? If we cut these hours by half next sprint, what does that buy us? Express it concretely: hours per engineer per sprint, calendar days reclaimed and additional stories delivered.

For automation, name the productivity multiplier — manual work is repeated every sprint. Cutting four hours of repeated manual work per week recovers ~200 hours per engineer per year.

For rework, frame it as a defect-curve cost — cheaper to prevent than to patch. The point isn’t precision; it’s making the opportunity tangible enough that the team feels the shape of the lever.

For planning, acknowledge it’s diffuse — a hundred small things rather than one big one — so the recovery framing is “compound effect across the sprint” rather than a single dramatic number.

For context switching, keep it brief: name the cost, note it as a multiplier on the other categories rather than as a standalone target.

Exercise 3 - Voting

This should be a quick exercise. The team can use up to 3 votes per team member to indicate what’s most important to them.

Start the voting cycle from the Workspace menu. You can use the 2:00 minute countdown timer to facilitate the exercise and get the team actively involved. (Don’t worry, if you are running a one-hour workshop timer, once the voting timer is done your workshop timer will resume).

While voting is enabled, everyone’s mouse pointer will change to a crosshair and any card they click on will consume one of their votes. They can undo a vote using the speed dial menu.

Exercise 4 — Inside control vs. inside influence (10 min)

Two “influence wells” on the board:

• Inside our control. Things the team can change without needing anybody else’s permission — acceptance criteria, code review depth, test coverage discipline, our own standup hygiene, how we capture decisions and which tools we adopt within our boundary.
• Inside our influence. Things the team can advocate for but doesn’t decide alone — cross-team handoff cadence, third-party tooling budgets, leadership decisions about meeting culture, hiring, deadlines and scope.

Walk the dominant categories from Exercise 2 and place each visible source of waste into one of the two columns. Do this by dragging the cards from the board into one of the influence wells on the left. Some will straddle — capture the most accurate rather than force-fitting.

We act on the control column tomorrow, and we build the case from the influence column.

The GembaKai readout is an evidence pack for both — actions for what we control, business case for what we influence. Cite Stephen Covey’s circle of influence vs. circle of concern if it helps the team — the language has been around since 1989. We split influence into “control” and “influence” here because at the team level the distinction matters: things you can change tomorrow versus things you have to argue for.

We drop the outer concern circle here — anything outside the team’s influence isn’t actionable from a one-hour readout, and leaving it on the board pulls attention away from the two columns that can produce moves.

The three waste-walk workshops in our suite — automation, rework and planning — each put both halves in the team’s hands. They produce concrete actions for the control column and a structured argument the team can take to leadership for the influence column.

The blue cards

When you initially entered this exercise, the system automatically created four blue cards for each category of waste. These cards represent the total amount of lost effort in each category. Optionally, you can drag these into an influence well — but, be aware that the total effort will be carried with it. It will inflate your recovery estimates (in the next exercise).

Exercise 5 — Workshop selection (8 min)

Take the dominant category from Exercise 2 and read it through Juran’s vital few vs. trivial many lens — the 1951 reading of Pareto that quality work has used ever since.

The principle: don’t run a workshop on every category. Run one for the category whose hours dominate. The hours that aren’t in that category are the trivial many this cycle — log them, leave them on the table and revisit next time.

The mapping:

If two categories are within ten percent of each other on hours, default to rework first, then automation, then planning. Rework upstream tends to dissolve waste downstream — fix-meetings, manual validation and blocked dependencies — so solving it first lifts the next readout’s planning and automation numbers without a separate workshop.

The full reasoning behind this sequencing is captured on the guides overview. The Pareto framing itself is a reusable technique — see Pareto analysis for the underlying method.

Watch for: teams gravitating toward the workshop they’re most comfortable with rather than the one the data points to. The facilitator’s job in this exercise is to keep the conversation tied to the readout’s hours, not to the team’s preference.

When no workshop is the right answer

If every category on the readout is small — for instance, a low-double-digit total of non-value-add hours across a full sprint — say so. Close with “we’re already lean; re-run the waste walk in N sprints and check whether anything has shifted.” Don’t fabricate a workshop just because we held a readout session. A team that’s already running tight is one of the readout’s possible findings.

Exercise 6 — Commitment and close (2 min)

Capture three things:

1. Workshop chosen. Name it: rework, automation or planning.
2. Date and facilitator. Pencil a slot — typically within the next sprint, before the data goes stale.
3. Pre-actions, if any. Some workshops have pre-homework (e.g., the rework workshop asks each engineer to bring 3–5 specific cases). Flag now so the team has runway.

Close with one sentence: “Next session, [date], [workshop] — and we picked it because [the dominant category].”

Export the summary

Use the Export menu item to download detailed information on the outcomes of the session (the download provides more detail than the summary action plan shown in the sidebar — it includes votes, time impact, and more). You can download the outcome in multiple formats, including Markdown and RTF.

Capturing results

The scribe produces a short readout within 24 hours — a few sentences, written the way someone would actually summarize the meeting in chat. Cover the dominant category and its hours, the workshop the team picked, when it runs and who’s running it, what landed in the control versus influence columns, and any pre-actions with their owners. Something like:

Sprint 14 GembaKai readout — held Wednesday afternoon. Rework dominated at 22 hours, about 47% of our non-value-add for the sprint, mostly from re-implementing the checkout flow after the demo. We’re running the rework workshop two weeks from Monday; Priya will facilitate.

On the control side we flagged tightening acceptance criteria and adding a code-review check against the AC. On the influence side we flagged the monthly demo cadence — the readout will go to product leadership as our argument for moving to bi-weekly.

Pre-actions: every engineer brings 3–5 specific rework cases to the workshop (owner: Priya, by next standup).

Post it to the team’s shared space and link it from the GembaKai readout so the choice has a paper trail. Six weeks from now the next readout will measure whether the chosen workshop moved the number.

Follow-up

• Within 48 hours: confirm the workshop date on calendars; circulate the chosen workshop’s pre-homework if any.
• Mid-sprint: facilitator pings the team to make sure pre-homework is happening.
• Next GembaKai readout: open by re-reading this session’s chosen category and asking “did the workshop move the number?” That single question is what closes the loop.

Alternative paths

Multiple categories tied

If two or three categories are genuinely co-dominant, sequence them across consecutive cycles — typically two-sprint intervals, so each workshop has time to land real change before the next readout. Pick the order using the rework-first heuristic.

Team is already lean

If the readout shows little non-value-add activity, the right move is no workshop this cycle. Re-run the waste walk in two or three sprints. Document the decision so it doesn’t look like the team forgot.

Business case for external advocacy

If the dominant waste sits firmly in the influence column from Exercise 3 — for instance, planning waste driven by mandated cross-team ceremonies the team can’t unilaterally cut — the readout itself is the artifact. Promote it (with the team’s permission) into a leadership conversation rather than holding a workshop for which the team can’t deliver actions. The relevant follow-on workshop will then be informed by leadership’s response, not by a session held in a vacuum.

References

Related guides

• Rework workshop — for tracing rework hours back to their origin.
• Automation workshop — for surfacing manual work the team has normalized.
• Planning workshop — for finding meeting overhead, late decisions and external dependency friction.

Process guide

• Pareto analysis — the 80/20 method behind workshop selection.
• Gemba walk — for going beyond the readout’s aggregate numbers and observing work in flight.
• 1-2-4-All — useful when the room is large enough that a single open discussion in Exercise 3 won’t surface every voice.
• Dot voting — fall back to this if Exercise 4 stalls on workshop selection and the team needs a quick convergence vote.

Delivery Playbook articles

• My formula for running a successful workshop — the meta-framework these guides follow.
• Using a Waste Walk to improve your “ways of working” — the full waste walk methodology this readout closes out.

External references

• Stephen R. Covey, The 7 Habits of Highly Effective People (Free Press, 1989) — the original circle of influence / circle of concern framing referenced in Exercise 3.
• Vilfredo Pareto, Cours d’économie politique (1896) — the original observation behind 80/20.
• Joseph Juran, Quality Control Handbook (McGraw-Hill, 1951) — Juran’s “vital few vs. trivial many” reading of Pareto, applied to quality work.
• Lean Enterprise Institute, Gemba walk — the canonical reference for going where the work happens.