A3 problem solving example: follow the evidence across the page
A quoting team reduces late replies from ten to four, but misses its target. A completed A3 shows the problem, the reasoning, the trial and what still needs work.
The improvement report says the new quotation checklist is live. Everyone has been trained. The action list is complete.
Customers are still waiting for quotes. To judge the change, the team needs to see what happened to those requests after the checklist went into use.
An A3 is a concise report that connects a problem, its investigation, proposed changes and follow-up. The name comes from the paper size. The Lean Enterprise Institute describes it as a way to guide problem solving and discussion, with an owner responsible for moving the work forward.
A reader should be able to follow why the team chose a change, what it expected and what it found. Filling every box won't supply a missing explanation.
Begin with a gap the team can check
Take a fictional supplier preparing routine customer quotes. It promises a reply within two working days of the first request. Time spent asking for missing details stays on that clock.
In a completed group of 30 eligible requests, ten quotes were late. Seven of those waited while the team clarified product dimensions. Three waited for price approval. The other 20 met the promise.
The team reviews the request messages and timestamps with sales and engineering. In the seven dimension-related cases, the quote could not be completed until the measurements were clarified. Several exchanges happened only after engineering picked up the request.
That gives the team a specific lead to test: ask for the needed dimensions during the first sales response, using a short illustrated guide. It hasn't established that every late quote has the same cause.
Put the whole A3 in view
The example below uses seven sections, with the reasoning on one side and the proposed work and follow-up on the other. Formats vary. The Managing to Learn excerpt describes similar elements and explicitly allows the layout to fit the problem.
Original A3 / fictional quote process
From the observed wait to the next decision
Owner: sales operations lead. The full story stays visible. Use the controls below to trace the evidence behind the review.
Background
Routine customers need a quote within two working days. A late reply makes their purchasing decision harder.
Current condition
30 requests: 10 late. Seven waited while dimensions were clarified; three waited for price approval. Clock starts at the first request.
Target
Next 30 eligible requests: at most 3 late and no incorrect quotes. Keep clarification time inside the same clock.
Analysis
Trace the seven dimension-related waits. Check which measurements were missing and why clarification started only after engineering picked up the request.
Proposed change
Sales sends an illustrated dimension guide in its first response. A named engineer helps resolve unanswered questions. Price and technical checks remain.
Trial plan
Sales operations lead owns the trial. Brief sales and engineering on day 1, use the guide from day 2, then review the next 30 eligible requests after all have final replies.
Follow-up
Pilot: 4 late, including 1 dimension wait and 3 approval waits. No incorrect quotes found. Target missed. Keep the pilot bounded, review remaining waits and agree the next test.
Scroll the outcome bars sideways to compare the two complete groups.
Four late quotes exceed the target of three
The late rate improves from 33.3% to 13.3%, but the predeclared target was at most 10%. Zero incorrect quotes meets the other numerical condition. Report the unmet target alongside the improvement.
| Group | Requests | On time | Late | Late rate | Incorrect |
|---|---|---|---|---|---|
| Baseline | 30 | 20 | 10 | 33.3% | 0 |
| Pilot | 30 | 26 | 4 | 13.3% | 0 |
The current condition leads to the proposed change: bring dimension clarification forward. The target tells the reviewer how much improvement the team is seeking. The follow-up records the actual result, including the remaining delay.
The owner in this example is the sales operations lead. Engineering and the pricing team need to help investigate and test the change; the name on the page doesn't give the owner authority over their schedules.
Separate what you saw from what you think it means
“People send bad requests” is too vague to guide a trial. The observations are more useful: which dimension was missing, when someone requested it and how long the quote waited for the reply.
The explanation still needs checking. Perhaps the form doesn't show which measurements matter. Perhaps customers need help finding them. Perhaps engineering receives the request before sales has had a chance to clarify it.
For the trial, sales sends the illustrated guide in its first response and flags unanswered dimension questions to a named engineer. The proposed mechanism is earlier clarification, while the quoted price and technical checks remain required.
The 5 Whys example shows how to test a sequence of explanations. An A3 can summarise that investigation and link to the supporting records rather than squeeze every email onto the page.
Compare the result with the target
The next 30 eligible requests produce four late quotes: one waiting for dimensions and three for approval. No incorrect quotes are found in the completed baseline or pilot groups.
| Measure | Baseline | Trial | Trial target |
|---|---|---|---|
| Requests followed to final reply | 30 | 30 | Next 30 eligible requests |
| Late quotes | 10 | 4 | At most 3 |
| Late quote rate | 33.3% | 13.3% | At most 10% |
| Incorrect quotes | 0 | 0 | 0 |
The late rate falls by 20 percentage points. The target is missed by one quote. The A3 should show both facts.
This before-and-after comparison supports another review. It doesn't isolate the checklist's effect from workload, customer mix or staff availability. The team should inspect the request histories and whether the guide was actually used before claiming the change caused the reduction.
The three approval delays also remain. They need their own explanation; adding more dimension questions won't necessarily help them.
A3 and PDCA describe connected work
PDCA means Plan, Do, Check, Act: plan a change, test it, compare the result and decide the next step. An A3 can hold the reasoning and evidence from that cycle in one readable view.
The PDCA worked example explains how a result can improve one measure and fail another. Our quoting example instead improves the chosen measure but falls short of its target.
For a manufacturing case and an existing printable worksheet, see Vrolen's A3 guide. It uses a label-alignment problem, separate from this office process.
Leave the next decision visible
For this trial, the team keeps the guide in the pilot group while it reviews the remaining dimension case and the approval route. It records the owner and review date rather than marking the whole problem solved.
If further testing supports a lasting change, update the working instructions and agree who watches performance after the trial. The standard work and SOP article covers that handover.
For your next A3 review, follow one proposed action back to the evidence that justified it. Then follow it forward to the measured result. If either connection is missing, the page needs more investigation, even if every task is ticked off.
Sources and example notes
- Lean Enterprise Institute: A3 Report. Definition, ownership and the role of discussion in problem solving.
- Lean Enterprise Institute: Managing to Learn excerpt, 2008, printed pages 7–11. Flexible formats, the usual A3 elements and their relationship to PDCA.
The supplier, observations, causes of recorded waits, targets and trial results are fictional. The seven-panel arrangement is an original teaching layout, not a required universal template. No causal effect, statistical confidence or business saving is estimated.