Why Systems Beat Memory

A system is a recurring task that has been written down so completely that a competent person who has never done it before could do it correctly, without asking anyone.

Here is the test that exposes everything:

For most small businesses, the honest answer is: almost everything. Prices are in the owner's head. Supplier phone numbers are in the owner's phone. The way to calm the difficult customer, the trick to starting the unreliable generator, the fabric quantity for a size-52 agbada, all memory.

This is exactly why investors walk away from profitable businesses (the very first insight of this manual): they are not buying the owner's memory, because the owner's memory is not for sale. A business that runs on memory is not a sellable asset. It only keeps working as long as that person keeps showing up.

Imagine a power cut at 9 a.m. A wedding order must leave by noon. The one person who knows how to start the generator is away for the day.

Everyone wants to help. One person turns the choke all the way. Another pulls the cord again and again. The generator floods. Work stops. The order goes out late.

Nobody was careless. The business simply kept an important instruction in one person's head.

This is the real cost of relying on memory:

• Work changes depending on who is present, so quality drifts.
• New staff must interrupt experienced staff to learn ordinary jobs.
• When someone leaves, useful know-how leaves too.
• Cash and approvals are easier to misuse when nobody has written down the right steps.
• The owner becomes the person everyone waits for.
• A buyer cannot take over a business that only works while its current people remember everything.

MANIAC MINDZ's old generator had a trick: choke halfway, wait three seconds, then start. If the choke was set wrongly, the generator flooded and would not start for an hour.

For two years, only the senior tailor knew the trick. The day he travelled for a funeral, power failed at 9 a.m. with a wedding order due at noon. Four people took turns yanking the cord. The generator flooded. The order went out four hours late.

The fix took one laminated card, zip-tied to the generator handle:

Cost: a few naira and five minutes.

Notice what the card changed. It gave the next capable person the right first steps. It included a recovery step if the generator flooded. It was attached to the machine, exactly where it was needed.

The generator did not become easier to start. The knowledge became available to the business, not just to one person.

A useful system does not need to be a long document. Use this simple test instead:

1. Put the instruction where the work happens.
2. Ask someone capable, who has not done that task before, to use it.
3. Watch for the first place they get stuck.
4. Add the missing step, then try again.

For the generator, the card passed because a capable person could find it, follow it, and recover from the common mistake without calling the senior tailor.

You can use different formats for different jobs:

Only now do the three useful labels matter:

• Findable: the person can locate the instruction at the moment they need it.
• Followable: the steps make sense to a capable person who did not write them.
• Current: the instruction still matches the way the work is done today.

The test is never "Is it long?" The test is: Could a capable person complete this task tonight using only this instruction?

1. Run the Two-Week Test on your business. List everything that would stop.
2. Beside each item, write whose memory it currently lives in.
3. Pick the single item where failure would hurt most this month.
4. Write its laminated card tonight, steps, plus the one "if it goes wrong" line. Put it at the point of use.
5. This week, have someone else do the task from the card without you present. Every question they ask is a missing line; add it.

You've just built your first system. Chapter 2 shows how to build them properly, every time.