There was one month where everything happened at once, and for about three weeks I let myself think a corner had been turned.

It had not been turned. But I want to write down what that month looked like from the inside, because it is the closest I have come, and the way it fell apart was more instructive than any of the individual failures on this site.

Three Things Landing in the Same Three Weeks

None of it was planned. That is the first thing worth saying — the month was not the result of a strategy, it was the result of coincidence, which is exactly why I should have been more suspicious of it than I was.

The first thing was a run of good delivery weekends. Two in a row where the weather and the timing cooperated, both meaningfully above the median, both of which I understood at the time to be luck and both of which I quietly incorporated into my expectations anyway.

The second was a beta testing payout that was genuinely larger than usual. Not large. Larger. The kind of thing that arrives once or twice a year and feels like finding money in a coat.

And the third was that the tool with two users briefly had, for about ten days, four users. I do not know who the other two were. I have never known. They came from somewhere, used it properly for a week and a half, and left.

The Spreadsheet I Built at the Peak of It

Week three is when I did the thing. I opened a spreadsheet and I extended the month forward.

I want to describe the specific error clearly, because it is the most common error in anything like this and I made it with a straight face: I took three weeks that were each independently at the top of their normal range, added them together, and then multiplied. Every input was a peak. The output was a number that had never happened and was not going to.

And it was not a wild number. That is what made it dangerous. It was a modest, believable figure that, if it had been real, would have been a genuine change in my household — not a different life, just a materially easier month, every month.

I looked at that spreadsheet a lot. I did not tell anybody about it, which in hindsight was the tell. I have described things at the dinner table on much weaker evidence. Some part of me knew that saying it out loud would break it.

Where It Stopped, Which Was Everywhere at Once

Week four the weather turned and delivery went back to the median. The payout did not repeat, because those payouts do not repeat, which I knew. The two extra users left without a word, and the tool went back to two, where it has been ever since.

Nothing failed. That is the part I keep coming back to. There was no disaster, no mistake, no thing I could have done differently. Every component simply returned to its ordinary level at approximately the same time, and the pile stopped looking like a trend and went back to looking like a pile.

The cost was about two weeks of being bad company. I was short with people. I was checking things. I re-ran that spreadsheet with lower assumptions maybe six times, trying to find a version where it still worked, which is the arithmetic equivalent of going back to the same parking lot.

"You've been in a mood," was the assessment at home, delivered as a statement of fact rather than a complaint, at which point I realized I had been in a mood about the ordinary continuation of my actual life.

The Rule About Multiplying, Which I Wrote on the Board

The rule is on the whiteboard in the garage in four words: never multiply a peak. Any month where several things go well at once is a month to enjoy and specifically not a month to extrapolate from, because the odds of independent good things repeating together are worse than the odds of any one of them repeating.

I also stopped keeping the running total. I used to have a figure in my head for what all the side stuff made together, and that figure was the thing that made a good month feel like a trend. Now each thing gets counted on its own and none of them are allowed to add up into a story.

Which sounds bleak written down, and is not. Three good weeks was three good weeks. I got to have them. What I did with them — build a spreadsheet and then be unpleasant to my family for a fortnight — was optional, and next time it will be optional again, and I will probably do it again anyway.

Everything went right for three weeks and then went back to normal, which is not a failure, and which took me about a month to stop treating as one.

Note: Everyday Hustle Jeff is the written side of a digital series. These accounts are composite and dramatized. Any figures are what they were on the day, for one person, and are not a projection, a track record, or advice.