The Same Improvement, Three Different Workshops? — A Case Study on Proposal Duplication and Knowledge Base Construction in an Electric Tool Company

By: QTank Published: 9/7/2026 Views: 85
Current rating: ★★★☆☆ Rate this Equivalent to 8 ratings

1. Three Out of Ten Proposals Are "Familiar Faces" in 4,800 Proposals

A certain electric tool manufacturing company has approximately 1,600 employees and eight workshops. Over the past five years, the company has promoted a proposal system for all employees, increasing the annual proposal volume from 600 to 4,800, with participation rates exceeding 50%. The "Improvement Stars" on the wall have been filling up, and the reports have been looking better each year. However, the quality department decided to take a "look back" at the proposals: they categorized the annual proposals by workshop, process, equipment model, and problem phenomena. The results were alarming—30% of the proposals were highly repetitive. The same improvement was proposed by Workshop 1, then by Workshop 3, and most likely by Workshop 5 next year. For example, someone proposed "excess stator coating," and three workshops wrote their own versions, using different wording but addressing the same issue.

A few examples were even more amusing. A veteran assembly worker with twelve years of experience carefully wrote a proposal to "prevent missing screws during model changeovers." During the review, it was discovered that a similar improvement had been implemented in Workshop 3 two years ago, even with a similar tooling structure. The old worker smiled bitterly: "If someone had told me two years ago, I would have preferred to spend that half-hour assembling two more machines." Another workshop proposed a protection plan for rotor end face dents, only to find out halfway through implementation that the neighboring workshop had already solved a similar problem last year, wasting a month of trial and error.

The quality department quickly calculated the threefold cost of duplicate proposals: first, the review cost, with engineers spending over forty hours each month reviewing old records and confirming old accounts; second, the incentive cost, where the same improvement was repeatedly rewarded, diluting the budget; third, the credibility cost, where employees felt that "proposals are useless and there's nowhere to check," gradually losing their enthusiasm. In the second quarter of the following year, the proportion of effective proposals dropped below 20%, and the system was once again at a crossroads.

2. The Root Cause: A "Mailbox" but No "Library"

During the post-mortem meeting, the promotion office summarized the issues into three points. First, proposals only had records but no knowledge—the review conclusions, implementation results, and effect data were scattered across various departments' spreadsheets and paper review forms, never compiled into a searchable archive. Second, there was no guidance at the submission end—employees wrote proposals based on their memory and observations at their workstations, unaware that similar problems had already been solved in other workshops, leading to inevitable repetition. Third, the results were not fed back—once a proposal was adopted and implemented, and the reward was distributed, no one documented the "how it was improved, the results, and the applicable conditions," making the improvements invisible and unusable by others.

The director of the promotion office used a metaphor: over five years, the system had collected 4,800 proposals, but it had only built a "mailbox" and not a "library." The mailbox only handled incoming proposals, with no storage or borrowing functions. The employees' knowledge was scattered, and the improvement results were static. No matter how many proposals were submitted, they were just repeatedly drawing water from the same well. The attendees agreed: repetition was not due to employee laziness but to the system's failure to provide employees with a "window into history."

3. Entry Control: Pre-submission Duplication Check to Block Repetition

The first reform focused on "storage." The promotion office led the effort, spending two months to enter all 4,800 proposals into a unified ledger, coded by workshop, process, equipment, and problem type, along with whether they were adopted, the implementation results, and the reasons for rejection. For historical states that were uncertain, they invited veteran reviewers to confirm during the monthly review meetings, preferring to be thorough rather than hasty. After the ledger was established, a simple search function was added to each workshop terminal: before submission, employees could input keywords or workstation numbers, and the system would automatically display similar proposals and their current status.

The rules were also adjusted: if a similar proposal had already been implemented, employees could submit a "transplant suggestion"—moving the solution from Workshop A to Workshop B, noting the differences in the site, and using a fast-track review process to get a conclusion within a week. If a similar proposal had been rejected, the system would directly show the rejection reason, helping employees avoid the pitfalls of previous attempts and focus on new problems. For example, an employee in the pneumatic line searched for "missing screws" and found that the tooling in Workshop 3 from two years ago was a manual fixture, unsuitable for their line. They then submitted a transplant suggestion for a pneumatic line-specific solution, which was reviewed and approved within three days and implemented within two weeks, earning them an additional reward—this time, the improvement did not take a detour.

Three months later, the duplicate submission rate dropped from 31% to 18%, and the review meetings transformed from "account-checking sessions" back to "new-proposal evaluation sessions." Employees also benefited: spending two minutes checking before submission increased their success rate and eliminated the frustration of having their proposals rejected as duplicates.

4. Exit Control: One-Page Repository to Make Results "Snowball"

Blocking duplicates was not enough; genuine results needed to be retained and disseminated. The new rules required that after a proposal was implemented for a month and its effects confirmed, the implementer must present the "problem, solution, effect, and applicable conditions" at the review meeting within seven days. The results, along with before-and-after photos, would be recorded in the results repository. If the conditions were not ripe for immediate promotion, the reasons would be documented and archived to prevent future trial and error. The results repository was maintained by a dedicated person from the promotion office, who would publish a "new arrivals list" each month, highlighting the most replicable improvements at the pre-shift meetings in each workshop.

The incentive system was also realigned. The company introduced a "reuse reward": the original proposer's reward was tied to the number of times their proposal was transplanted, with an additional reward (200 to 500 yuan) for each adoption by another workshop, and a "most reusable proposal" award at the end of the year. The incentive logic shifted from "reward once for each proposal" to "reward once for each use," encouraging employees to write proposals that were clear and easy to transplant—detailed applicable conditions, investment costs, and key points. Over six months, more than forty proposals were transplanted three or more times, with one tooling improvement for preventing rotor dents adopted by four workshops, saving nearly 200,000 yuan in repeated trial and error costs.

5. Eighteen Months Later: Duplicate Rate Down, Benefits "Replicated"

Eighteen months later, the data confirmed the direction: the duplicate proposal rate dropped from 31% to 9%; the time engineers spent on "account-checking" was reduced to less than ten hours per month; the effective adoption rate increased from 21% to 38%; and over 130 cross-workshop transplants were recorded, with confirmed annual benefits exceeding 3 million yuan, nearly 60% of which came from "replication" rather than "original" improvements. The employees' perception was even more valuable: they could check before submitting, receive feedback after submission, and see their improvements used by others. Participation rates not only did not decline but increased, with employees actively applying to "claim" improvements from the results repository.

In retrospect, the lessons from this company are worth noting for every proposal system manager: a proposal system has two engines. One ensures that proposals are submitted, driven by incentives; the other ensures that proposals are retained and usable, driven by knowledge. Most companies focus all their efforts on the first engine, continuously increasing bonuses, but neglect the second engine—the duplicate rate is a thermometer for the system's health. A few practices can be replicated: build a ledger before discussing a system, and manually enter data if it is incomplete; use "pre-submission guidance" instead of "post-submission rejection" to reduce employee frustration; incorporate "reuse" into the reward rules to share the benefits of replication; and include the duplicate rate in the monthly report to make the problem visible. Improvements are not afraid of being small, but of being scattered; they are not afraid of being slow, but of being forgotten. Instead of rushing to increase bonuses, it is better to first build a "library" so that every improvement becomes a stepping stone for future efforts.


The upper limit of a proposal system is not in the mailbox but in the knowledge base: pre-submission duplication checks block repetition, post-implementation archiving preserves results, and reuse amplifies benefits, transforming improvements from "quantity stacking" to "snowballing."

Knowledge code: 5.1.2

Version: v20260907

Author: QTank QTank is dedicated to providing systematic professional knowledge, methodologies, and practical tools for quality management practitioners, helping companies continuously improve their quality capabilities.