Building an Employee Suggestion System

An employee suggestion system is a structured way for frontline workers to submit improvement ideas, get a fast response, see what happens to each one, and share in the results. The systems that work are not suggestion boxes: they close the loop quickly and visibly, so the people closest to the work keep contributing ideas.

The eighth waste of lean is non-utilized talent, one of the eight forms of muda and a suggestion system is the main way a plant stops wasting it. The operator running a machine 2,000 hours a year sees problems and fixes no engineer will ever notice. The question is never whether people have ideas; it is whether your system makes it worth their time to share one. This post covers why most suggestion boxes fail, a seven-step design that gets used, how fast to respond, what to measure, and how it feeds the wider lean and kaizen effort. It is a core piece of employee engagement in manufacturing.

What is an employee suggestion system?

An employee suggestion system is a defined process for collecting, evaluating, implementing, and recognizing improvement ideas from employees. At minimum it needs four parts: a way in (how ideas are captured), a way through (who reviews and decides), a way back (how the submitter hears the result), and a way to act (how good ideas actually get implemented). A suggestion box on the wall has the first part and none of the other three, which is why boxes fill with dust.

The difference between a program that averages a handful of ideas a year and one that averages dozens per person is almost never the people. It is the response. When an idea disappears into a box and nothing comes back, people stop submitting within a few weeks. When an idea gets a real answer in days and the submitter can watch it move, submissions compound. The system is a habit loop, and the loop only holds if it closes.

Why do most suggestion boxes fail?

Most suggestion boxes fail because they are open at one end and closed at the other: ideas go in, and nothing comes back. Four failure modes account for nearly all of it. First, no response, so submitters conclude the box is a black hole. Second, slow response, where an idea gets answered months later after the moment has passed. Third, no visible action, so even accepted ideas never seem to change anything. Fourth, judging ideas only by dollar savings, which teaches people that small quality-of-work-life fixes are unwelcome, when those are exactly the ideas that build the habit.

There is also a trust dimension. If the first honest suggestion about a real problem is met with defensiveness, the flow stops immediately. A working system treats a suggestion as a gift of attention, answers it straight, and never punishes the messenger. That posture is a leadership behavior, not a software feature.

How do you design a suggestion system that gets used?

Design the system as a fast, visible loop rather than a collection point. These seven steps build one that keeps producing ideas.

1. Make submitting frictionless. Put the way in where the work happens, on a tablet at the line or a simple card at the station, not a form in an office. If submitting an idea takes longer than the idea took to have, most people will not bother.
2. Acknowledge every idea fast. Confirm receipt within a day or two, by name. The acknowledgment matters as much as the eventual decision, because it proves a human read it.
3. Set a decision deadline. Commit to a yes, no, or "here is what we need to decide" within a fixed window, often one to two weeks. A prompt no with a reason beats a slow maybe every time.
4. Push decisions to the lowest level that can make them. A supervisor who can approve and fund a small fix on the spot keeps the loop fast. Routing every idea to a monthly committee is how programs die.
5. Make status visible to everyone. A board or screen showing each idea and its stage, from submitted to done, turns the program from a private transaction into a shared, contagious activity.
6. Implement quickly and give credit by name. Nothing recruits the next idea like seeing the last one become real with the submitter's name on it. Recognition can be small; visibility matters more than the size of the reward.
7. Close the loop out loud. When an idea is done, announce it and its result. Even a "we tried it, here is what we learned" closes the loop honestly and keeps trust intact.

How fast should you respond to a suggestion?

Respond within days, not weeks, because response speed is the single biggest predictor of whether the program survives. Acknowledge receipt within a day or two and reach a decision within a week or two for routine ideas. The reason speed matters so much is that a suggestion is a small act of engagement, and engagement is fragile: the submitter is watching to learn whether their effort mattered. A fast, plain answer, even a no, tells them it did. A three-month silence tells them it did not, and that lesson spreads faster than any poster encouraging participation.

What metrics tell you the system is working?

Four metrics tell you whether a suggestion system is healthy, and none of them is dollars saved. Participation rate is the share of employees who submit at least one idea in a period; a rising participation rate is the clearest sign the loop is trusted. Ideas per employee measures depth of engagement. Implementation rate the share of submitted ideas that get acted on, tells you whether the system produces change or just collects it. Response cycle time from submission to decision, is the leading indicator; when it creeps up, participation falls a few weeks later.

Benchmarks vary widely and should be treated as direction, not targets. Mature Japanese programs are famous for very high participation and dozens of ideas per person per year, while many Western suggestion-box programs historically saw only a small percentage of ideas accepted. The gap is not cultural destiny; it is the difference between a closed, fast loop and an open-ended box. Chase your own trend upward rather than a headline number from another plant.

How does a suggestion system connect to kaizen and continuous improvement?

A suggestion system is the everyday, distributed form of kaizen. Where a kaizen event concentrates a cross-functional team on one problem for a focused week, a suggestion system captures the steady stream of small improvements between events, from the people who see the process every shift. The two reinforce each other: events surface ideas that feed the system, and the system surfaces recurring problems worth a full event, the kind a fishbone diagram or a formal root-cause session can then take apart. Both depend on the same shop-floor habits that a gemba walk builds, where leaders go to the work, ask what gets in the way, and act on the answers.

It also feeds the wider knowledge base of the plant. Many of the best suggestions are really tribal knowledge being written down for the first time, the fix an experienced operator has done for years but never documented. Capturing that in a system, rather than losing it when the person retires, is one of the quieter but larger payoffs.

How do you keep a suggestion system alive long term?

Programs decay the same way they fail: the loop slows, then stops. Sustaining one means watching response cycle time like a vital sign and treating a rising number as an emergency, keeping status visible so the activity stays social, and making sure decisions still happen at the lowest capable level rather than drifting up to a committee. This is where doing it on paper works against you, because paper cards get lost, status is invisible, and no one can see the response time creeping up until participation has already collapsed. Capturing ideas, routing, decisions, and status in one place that everyone can see keeps the loop fast and honest. That kind of real-time, visible operational system is what CLS built with Harmony (see the CLS case study), turning scattered floor knowledge into something the whole plant can act on. No rip-and-replace, just a loop that stays closed.