The suggestion box nobody empties: an engagement process that actually closes the loop

A business of a hundred and forty people had a suggestion box by the canteen, a feedback form on the intranet, a WhatsApp group the founder had started and an annual survey. Somebody counted, out of curiosity, and found forty-one things had been raised across the four routes in a year. Nine had been acknowledged. Three had produced a change anybody could point to. Nobody had ever told the people who raised the other thirty-eight what happened, because nothing had.

The business did not have an engagement problem in the way it thought. It had four collection points and no process — and collection without process is worse than silence, because it teaches people that raising something is a way of being ignored in writing.

Most engagement effort goes into the front end: the survey design, the form, the box. That is the easy part. The process is what happens in the two weeks after something is raised — who sees it, who decides, who replies, and how the person who raised it finds out. Businesses that skip that part have built an inbox, not a process.

The test is simple and usually uncomfortable. Take the last ten things employees raised, through any route, and answer three questions for each: who owned it, when was it answered, and does the person who raised it know the outcome. If the honest answer to the third is mostly no, the box is decoration.

1  RAISED      through a route people actually use
2  LOGGED      one place, one number, one owner named
3  ANSWERED    within a stated time - even if the
               answer is 'not yet' or 'no, because'
4  ACTED ON    or explicitly declined with a reason
5  CLOSED      the raiser is told, in writing

   Nothing moves to the next stage silently.

Stage three is where most processes die. An answer is not an action; it is an acknowledgement that a human read the thing and a date by which something will happen. That alone changes how people experience raising something, because it converts a void into a conversation. Most businesses can commit to answering within five working days. Very few do, because nobody was ever asked to.

Four routes feel generous and behave badly. Each has a different audience, a different owner and a different chance of being read, so the same concern raised twice gets two different fates and people learn to game the routes — or, more often, to stop. Consolidating to one route with one log is the single most useful structural change, and it is resisted because closing the founder's WhatsApp group feels like closing a door. It is the opposite: it is the first time a door has led anywhere.

• One log. Every item, whatever route it came in by, gets a number and an owner in the same place. A spreadsheet is fine to start.
• One owner per item. Not HR by default. The manager or function that can actually decide. HR runs the log, not every answer.
• A published response time. Five working days to a first answer is achievable and credible. Silence past that date is visible on the log.
• Declines are answers. 'We are not doing this, and here is why' closes an item honestly. Leaving it open forever does not.
• Say what changed, out loud. Two items acted on and announced do more than twenty absorbed quietly.

A single route will receive things that are not suggestions at all: a complaint about a colleague, an allegation of unfair treatment, a safety concern. These must not sit in the same queue as a request for a better chair, and a process that treats them identically will either bury something serious under the five-day response time or frighten people off raising anything.

The fix is a triage step at logging, done by whoever runs the log, with three outcomes: this is a suggestion and follows the loop; this is a grievance and goes to the grievance and disciplinary process immediately, with the raiser told that is where it has gone; this is a safety matter and goes to whoever owns that, today. The routing rule is written down so it does not depend on the judgement of whoever is on duty.

An employee engagement process that has this triage step built in is one people trust with the small things, because they have seen that the serious things are handled differently. One that does not is one people learn to keep the serious things away from.

The loop breaks most often at the manager, because most items end up with one and managers were never told this is part of the job. A manager who receives an item, agrees it is reasonable and does nothing has not been lazy — they have no process telling them a reply is expected by Friday. Making the response time a stated expectation of every manager, with the log visible to their own manager, is what makes the loop hold. It is a small change and it is the one that matters.

This is where the engagement process and employee engagement and culture work meet: the process is the mechanism, and the culture is whether managers believe it applies to them.

Sentiment scores tell you how people feel about the loop; they do not tell you whether it works. Measure the loop itself, and the numbers are more useful and much harder to argue with.

1. 1

   Items raised per month

   A rising count after the process is introduced is good news. It means people believe raising something leads somewhere. A falling count is the earliest sign the loop has broken again.

2. 2

   Days to first answer

   Median and worst case. The worst case is the number to watch; one item left for six weeks undoes the credibility the other twenty built.

3. 3

   Proportion closed with a stated outcome

   Acted on or declined with a reason. Items that simply age out are the failure mode, and this number makes them visible.

4. 4

   Raiser told

   Yes or no, per item. It is the cheapest field to track and the one that most determines whether anyone raises the next thing.

Where the same category of item keeps arriving — rosters, a piece of equipment, one team's workload — the loop has found something the process cannot fix on its own. That is a finding for the HR process audit rather than another reply, and the log is what makes it visible.

A business that has collected feedback for years and answered little has a credibility deficit, and a new process launched over it will be met with the scepticism it has earned. The honest way through is to say so: acknowledge that things were raised and not answered, publish the new route and the response time, and then hit the response time without exception for three months. People will test it. Two or three visible closures in the first month are worth more than any announcement.

The design work — one route, one log, named owners, a stated response time and a way of telling people what happened — is what an employee engagement process engagement builds, and it is deliberately unglamorous. The suggestion box was never the problem. Nobody emptying it was.