Skip to content
Blogs Best-practices

Process improvement after three months: why the old way returns

Process improvement after three months: why the old way returns

A process improvement almost always starts well. The new system is ready, the team is enthusiastic and the first week feels like a fresh start. Three months later, half the team is back in Excel and the rest have slipped back into the old way of working.

As a business owner, you know the reflex that follows. The system is no good, the software is too complicated, we should have chosen a different package. It is almost never the system. It is how it was introduced. And that is unwelcome news, because buying a different package is easier than admitting that the rollout was the weak spot. Anyone who switches to new software after three months usually repeats exactly the same mistakes, only now with a new invoice and a team that is even less keen on the next attempt.

The first mistake is already in the setup. The new process is built perfectly in one go: every field, every automation, everything that might ever come in handy. For the person building it, that feels like thorough work. For the people who have to work with it every day, it is too much. At one of our clients, at some point 80% of the fields were simply left empty. Nobody had decided to skip them. They were quietly ignored, because nobody understood why they were there or what happened to the information entered.

On top of that come two patterns you only see after a few weeks. The process depends on one person who has to drive it, and exceptions creep in. We will just handle this one client the old way for now. That sounds harmless in week two. By around week eight, that exception has become the new rule. So a process improvement rarely fails on the day you introduce it. It fails in the weeks that follow, when nobody is keeping watch.

The rollout is the real work

Businesses that are still sticking with it after three months usually do not have a better system. They treat the weeks after go-live as the most important part of the project. The setup is the beginning. Usage in weeks one to twelve decides whether it lasts. That calls for a different order than most business owners are used to, and above all more patience with automations.

They start simple. First the flow, so everyone understands how the work runs from start to finish. Then practice, with real work. Only once the process is being used do the automations come in. That order feels slow to anyone who already knows exactly what the system can do. But an automation on a process nobody uses automates nothing. It only raises the barrier for the people who were already in doubt.

They also agree in advance when the improvement counts as a success, and look at that every week at a fixed time. No big evaluation report after six months. A short check, so you can adjust course before an exception becomes a habit.

What does such a check look like in practice? Fifteen minutes on a monday morning is enough. You look at three things: are the most important fields being filled in, are all ongoing jobs in the system, and have any lists popped up outside the system. You do not need to build a dashboard. A few spot checks often say more than a report. What matters is that the slot is fixed in the calendar and does not get dropped as soon as things get busy, because it is precisely in busy weeks that people fall back on what feels familiar. Anyone who skips that check twice usually only notices something again when it is too late.

Four failure points, four countermoves

If you put the patterns side by side, you see that each failure point has a direct countermove. Too much at once versus starting simple. One driver versus bringing the whole team along. Exceptions versus a single way of working. No monitoring versus a weekly check-in. It helps to see them this way, because then for every rollout you can check which of the four you have not yet covered. Usually two are missing, and they are almost always the last two, because they only start to count after go-live.

Why one driver is not enough

The usual approach is to appoint one person and have them trained, in the hope that they pass it on properly to the rest. If that person gets stuck, becomes frustrated or leaves, everything falls over. Your old process ran on the memory of a few people. Your new process then runs on one person. You have moved the risk, not removed it.

Businesses that stick with it have everyone go through the process themselves, with a real project. Not a demo with made-up data, but a job that was already on the schedule this month. That way every team member discovers for themselves where it chafes, and the questions surface in week one instead of week ten.

Why exceptions are so costly

Every exception is a step backwards. That sounds strict, and it is. As soon as a second list exists alongside the system, everyone has to remember which client is where. After a few weeks the team is unsure which source is correct, and then the old, familiar Excel file wins. The businesses that stick with it are consistent about this: the new way is the only way, even for that one difficult client.

The weekly check-in makes that achievable. If you see in week three that fields are staying empty or that someone is slipping back, you are talking about one client or one colleague. If you only notice after three months, you are talking about a whole team that went back to the old habit long ago, and you are effectively starting over.

A process improvement stands or falls with the twelve weeks after launch. Start simple, bring the whole team along, allow no exceptions and check every week at a fixed time how it is going. Those who do that are still working the new way after three months. Those who skip it will be comparing software again in a quarter's time. Lead Innovation helps business owners set up those first weeks properly, so the new process holds once the enthusiasm of the first week has worn off.

Want to know more about monday.com?

The Lead Innovation team at work