Nothing is on fire. The Monday report still goes out. The team is still hitting most of its numbers. The founder is busy enough not to feel idle, and tired enough to feel productive. From the outside, the business looks fine. This is exactly the window in which drift is doing its work.
Drift is the Monday report landing on Tuesday morning, then Tuesday afternoon, then sometimes Wednesday, and nobody pushing back because everybody is busy. It is the same question answered two different ways by two people on the same team in the same week, and neither of them noticing because they were not in the same meeting. It is a decision sitting in a queue for nine days because the one person who can sign it off is on holiday, and nobody else has the authority or the context to move it on.
It is a junior asking why a process is done a certain way, and the senior saying yeah we just always do it that way, and neither of them stopping to write down the answer. None of that is broken. All of it is drift. By the time it breaks, it has been drifting for six months, and the fix is no longer a fix, it is a clean-up.
What to look at when you audit.
Pick a Tuesday afternoon. Close the laptop, open a notebook, and walk the floor. The whole method is three questions, asked honestly, written down, and revisited at the end of the week. You are not building a deck. You are not hiring anyone. You are just looking at the parts of the business you have stopped looking at because they have not screamed for attention.
Question one: report cadence. Pick the three reports your business actually runs on (weekly numbers, cash position, ops dashboard, whatever yours are). For each one, write down the day and the time it landed last week. Then the week before. Then the week before that. If the landing slot is moving later, or moving around, that is drift. The report is the thermometer. It is not the disease, but it tells you the disease is there.
Question two: the decision queue. Open every channel where decisions wait on you (email, WhatsApp, the team chat, the approvals folder, your own head). List everything older than seven days. For each item, ask who else in the business could clear it if you were on a plane for two weeks. If the honest answer is nobody, that item is drift wearing a tie.
Question three: the knowledge gradient. If your most experienced ops person took two weeks off tomorrow, what would slip first? Write the names of the three things, then write the name of the person who would actually catch them. If the second column is blank, you have just found your audit.
What to do with what you find.
Most audits surface three or four things. The temptation is to feel embarrassed, write a plan that fixes all of them, and announce it to the team on Monday morning. Do not do this. The reason the drift accumulated is the same reason the all-of-it fix will fail; nobody in the business has the attention budget to absorb four operating changes at once on top of the actual work.
Pick one. The smallest one with the highest return, usually. Build a system around it, not a tool. The Monday report lands late because the data has to be assembled by hand on Sunday night; the system is to move the assembly to a workflow somebody owns, not to buy a new dashboard product. The decision queue is full because three of your team need a permission they do not have; the system is to grant the permission and write down what it covers, not to schedule a weekly approvals meeting. Build the system. Leave the others on a list.
Three months from now, run the audit again. Same notebook, same three questions, a fresh Tuesday afternoon. The first item is closed. Pick the next one. This is the part most businesses get wrong; they treat the audit as a one-off rescue mission, when the value is in the cadence. Audits are cumulative. You do not have to fix everything in one quarter. You have to keep looking, and keep picking the next one, and you have to be the person who looks.
Everything in this method is free. There is no software to buy, no consultant to hire, no quarter-long programme to launch. The founder does it herself, on a Tuesday afternoon, with a notebook, and the entire investment is four to six hours of attention spread over a week. The cheapest information your business can buy is the information you write down yourself, walking the floor, looking at the parts of the business that have been quiet for too long.
The choice on the table is simple. You can keep working, head down, until the morning a number does not match and the team has lost the thread, and then spend the next quarter cleaning up what should have been a one-afternoon fix. Or you can spend the Tuesday. Find the drift while it is still drift. Pick one thing. Build the system around it. Run the audit again in three months. That is the work. The rest is just waiting for something to break.