Business improvement proposal
Connect the problem and evidence to a proposed change, expected benefit, resources, implementation, and review.
Use a proposal when an improvement needs someone's decision or resources. Make the request checkable: what is happening, what you propose, why it may help, what it would require, and how a pilot would be evaluated.
Evidence of the problem, verified explanations and remaining hypotheses, an improvement option, estimated resources, and the decision or approval needed.
An editable seven-part proposal that separates evidence, expected benefit, resource estimates, implementation responsibilities, and risk checks.
How to use it
- State the problem with its scope, baseline, sample size, and measurement period. Mark missing information rather than filling it in.
- Distinguish a verified explanation from a hypothesis. Explain how the proposed change would address the observed problem.
- Define the expected benefit and how it would be measured. Keep the target separate from historical and actual pilot results.
- Estimate people, tools, budget, and opportunity cost. Clarify the total effort, its allocation, and what has been committed.
- Describe the implementation owner, pilot scope, schedule, dependencies, and approval request.
- Agree risks, review measures, and conditions for stopping, adjusting, or expanding before starting the pilot.
Keep its limits in mind
A written proposal does not grant approval or establish a return on investment. Targets and estimated effort remain assumptions until checked. Staff effort in person-days is not elapsed calendar time or a cash budget. This tool organizes supplied facts without calculating feasibility, allocating resources, or approving a plan.
Request a limited purchasing-form pilot
Fictional example: Xinggang wants to standardize internal purchasing requests. Eight of 20 sampled requests were returned for missing information, often budget allocation or specifications. The sampling period is not provided. A department manager's approval is needed; no approval or pilot results are supplied.
On a small screen, scroll across the diagram to read it.
Baseline and target: Eight out of 20 is a historical 40% return rate. The 20% target would be a decrease of 20 percentage points, not a measured pilot outcome.
Resource request: Configuration and training are estimated at two person-days in total. The two-week schedule is elapsed time, not two weeks of full-time effort. Neither the allocation between people nor a cash budget is supplied.
Decision needed: The purchasing team is the proposed implementation owner. Approval and specific stop or adjustment criteria still need confirmation before the one-department pilot.
Next step: Ask the purchasing owner to confirm the fields, and obtain the department manager's decision on the limited pilot with agreed review criteria. Record actual results before requesting wider use.
Three reasons to pause before requesting an improvement
Begin with the decision the recipient can make. State the affected work, the evidence of a problem and the bounded change you want to test. A proposal should let a manager question the assumptions and authorize a specific scope, rather than simply agree that the situation deserves attention.
Pause when the observed symptom has been renamed as a proven cause. Pause when an estimated saving is presented as an achieved result or a cash return. Pause when the requested people, authority or operating conditions are missing. A complete-looking form cannot supply those facts. Label the gaps and ask the relevant owner before promising implementation.
Separate the case for investigating a change from the case for adopting it. Keep the current process, no-change option and any smaller alternative visible. A pilot can reveal that a new form adds burden without fixing the first missing input. Make room for that outcome in the proposal.
Choose a request, a redesign or a test record
This is a site-defined proposal worksheet. It organizes a decision request; it is not a prescribed approval standard. Other methods may supply inputs, but they do not replace the decision owner’s actual response.
| Method | Starting material | Output |
|---|---|---|
| Improvement proposal | Problem evidence, candidate, resources and decision scope | A request with conditions and review |
| ECRS | Current steps and necessary outputs or controls | Redesign candidates to assess |
| PDCA | Agreed test plan and execution records | A comparison and next-action proposal |
| PREP | One recommendation with reason and example | A concise opening to the fuller request |
Worked case: distinguish the target from the observed return rate
This continues fictional Xinggang’s purchasing-form proposal. Eight of twenty sampled requests were returned for missing information, often budget allocation or specifications. The sampling period is not supplied. The historical sample rate is 8/20 = 40%. It establishes a problem to inspect, not that prompts will prevent every return or that all departments have that rate.
The candidate is one submission entry point with prompts for required information, using the existing form tool. The proposed return-rate target is 20%. Compared with the sample’s 40%, that would be a reduction of 20 percentage points, or 50% relative, if comparable observations supported it. It is not a measured reduction. Do not promise that a future batch will have twenty requests merely because the historical sample did.
The purchasing team proposes one week to configure and one week to pilot in one department. Configuration and training together are estimated at two person-days. The estimate supplies neither hours per person-day nor allocation between people, so do not convert it into a detailed roster or currency. The two-week window is elapsed planning time; the manager has not approved the pilot.
| Element | Given or proposed | Check before proceeding |
|---|---|---|
| Problem | 8/20 returned; period unknown | Retrieve scope, reasons and sampling conditions |
| Candidate | One entry with required-information prompts | Purchasing owner confirms fields and controls |
| Target | 20% returned, not achieved | Agree the same return definition and window |
| Resources | 2 person-days total; existing tool | Confirm availability, allocation and any cash cost |
| Pilot | One configuration week + one trial week | Manager decides scope; agree stop criteria |
Review the actual decision before reviewing the pilot
The conclusion is a bounded request with unresolved evaluation and resource details. Ask the purchasing owner to check fields, including whether every requested item is necessary. Ask the manager to decide the limited scope and record approved conditions, rejected conditions or a request for more information. A positive comment about the idea is not pilot authorization.
Before execution, define eligibility, the event counted as a return, treatment of repeated returns, observation completeness and a way to record filling time. Agree what would trigger adjustment or stopping; the case supplies no such thresholds. Keep unresolved and withdrawn requests visible instead of counting only convenient completed examples.
If the pilot occurs, retain actual effort, setup and training, filling time, return reasons and exceptions. Compare like-for-like outcomes and explain changes in volume or case mix. A lower return rate with much longer filling time may require redesign. Preserve the original estimate and target, append actual results, then make a separate request for broader use. The written proposal itself grants neither resources nor a return on investment.
Check your work
- Are the baseline, sample, period, and unknowns visible?
- Is the proposed change linked to evidence rather than an unverified root-cause claim?
- Are target benefits clearly separated from achieved results?
- Are total person-days, calendar duration, and cash budget distinguished?
- Are pilot scope, ownership, approval status, and evaluation conditions explicit?
Put the framework to work
Copy the outline and add your question, evidence, and open questions.
Practice with an example
Choose a preparation route for progress, a resource or decision request, or a period review. Use one service case to inspect openings, review checks and focused guides, and keep a preparation note.
Prepare a work update: start with the purpose →