The moment you decide what to fix in your business feels like the end of the story. According to the booklet, it is only the middle. A fix you did not check after it was carried out has proved nothing, and a fix unrelated to the cause you found repairs something else.
The failure that sounds trivial
The source names the failure outright, and it names it about the very people whose whole job is correction: sometimes a correction is carried out and nobody makes sure that the thing really got corrected.
It sounds obvious, and for that reason, says the booklet, it happens constantly. After you have found the problem and decided what to do, you feel finished. The problem is handled, and you can move on. But at that point nobody has seen whether anything changed.
Suppose an owner announces a new procedure at a team meeting. Two months later, nobody on the team has heard of it.
Now picture the same owner after she checks a few real cases two weeks after the announcement. Most of them did not follow the procedure, so the matter goes back on the table.
It is the same procedure both times. What changed is that the second time, somebody checked.
An announcement in a meeting passes on knowledge and promises nothing about doing. The whole booklet rests on that distinction: what a team knows and what it does are separate things, and most businesses test only the knowing.
A fix tied to the cause
The source gives two rules for verifying. The first is about the check itself: verify the data by looking. That often reveals the major problem that leads to the cause. Often what you are looking for is something omitted, so you need to know how the normal situation should look.
The second is about the correction: “Handle or recommend handling so that it stays handled.” The handling, the source adds, must relate directly to the cause that was found.
That gives two conditions. The fix has to hold over time, and it has to be tied to the cause. A fix that is not tied to the cause you found deals with a different problem, even when it looks sensible.
Suppose an owner finds that the delay is in approvals, and adds automatic reminders. The reminders have nothing to do with the cause, so they change nothing.
Now picture the same owner moving the approval authority to someone else. This time the fix is tied to the cause, and it holds.
According to the booklet, most fixes that fail fail in exactly this way. They are perfectly reasonable, and they fix a different problem from the one the diagnosis named.
Checking again is also how you find out that the cause itself was wrong. Suppose a garage owner sees fewer jobs, finds that nobody answers the telephone at lunchtime, and fixes it. Only when he checks again three weeks later does he see that the number has not moved and that the cause was something else.
First make sure there is a problem
The source adds a step that comes before everything else, and most of us skip it. First verify that a situation really exists and that you know what it is. The source puts it as a warning: don't go looking for the cause of a situation that does not exist.
Before you investigate, make sure there is something to investigate. A bad situation, according to the source, is measured by the difference between the existing scene and an ideal scene. And in a small business, says the booklet, a good share of investigations are into ordinary variation.
Suppose an owner investigates for a whole week why one month was weak. On the annual chart, that month turns out to be inside his normal range.
Now picture the same owner after he compares each month with the twelve months before it. Only two months out of twelve are genuinely out of range.
Why quality comes before marketing
The booklet ends on a statement that puts everything in order. According to the source, the goal an organization puts first is to deliver the highest technical quality it can in its field. In simpler words from the same source: “Quality is the only thing that counts.”
The booklet reads this as an order of operations. Before you increase what comes into the business, make sure that what leaves it is worth having. Marketing a broken product speeds up the damage.
Suppose an owner invests in a campaign while his delivery times have doubled. In practice he buys himself more disappointed customers, and disappointed customers usually stay silent and stop coming.
Now picture the same owner after he fixes the delivery times first. Six months later, the same budget brings in customers who stay.
The booklet gives this chapter two closing exercises, called verify and order. The practice section at the end also has one called the verification, which belongs to the routine.
The next time you fix something in your business, mark it done on the day you saw it hold.