You get to the review and the number has not moved. You have run the plan for a month, and the measure it was supposed to shift is sitting exactly where it was. Both available responses feel wrong. Changing the plan this early feels like flinching. Holding it feels like ignoring the one piece of evidence you went to the trouble of collecting.
Most people resolve this by doing neither. They carry on, and they review a little less often. The decision never gets made, it gets avoided, and the plan runs for another six months with nothing checking it.
The first question is not whether to change
A flat number is only evidence against a plan if the plan was supposed to be moving that number now.
My weight has plateaued recently, and my objective for the year has it lower than it is. Sitting in the review, the honest read looked like a plan that was not working. It was not that. My training through that period was pointed at performance rather than weight loss. The plan was doing what I built it to do. The measure I was anxious about was not the one it was aimed at.
That is a different situation from a plan aimed squarely at a number it is failing to move, and it deserves a different response. Both look identical in the review. A flat line and a feeling that something is wrong.
So the first question is not what to change. It is what this plan was supposed to move, in this period, and whether it did. If the answer is that the plan was never pointed at this number yet, you have not learned anything about the plan. You have learned that you are worried, which is not the same thing.
Most plan changes come from somewhere else
The clean version of this decision assumes plans get changed because an indicator goes flat. In my experience that is rarely what actually happens.
The clearest plan change I have made had nothing to do with a measure. The plan was to rebuild the platform, and to hold the existing one stable with as little change and attention as possible. The objective behind it was right and stayed right. What changed was that keeping the legacy system at arm's length turned out not to be viable, for reliability reasons and commercial ones. We needed to work on it actively, stabilise it, and add some smaller things the commercial team needed. The plan changed because an assumption inside it was wrong.
These are two different triggers and they do not behave the same way. One is a measure telling you the plan is not producing. The other is the world telling you the plan was built on something untrue. The second arrives faster, feels obvious when it lands, and tends to get acted on. The first is slower and quieter, and it is the one people defer, because a flat number always comes with a plausible reason to wait one more month.
A shape rather than a threshold
The tidy fix is to decide in advance what counts as failure. Set the number the measure has to hit by week six, and let the review be arithmetic.
I do not do this, and I do not think it works for most measures. Setting weekly or monthly targets on a lagging measure is restrictive and usually arbitrary. I set the year, break it into quarters, and put target weeks on actions because actions have end dates. The measures themselves I leave alone, because a month is a unit of my calendar and not a unit of anything that measure cares about.
What you can decide in advance is the shape. Some measures move incrementally and should show something small every week. Some move in steps, flat for a stretch and then jumping when something lands. Some do nothing at all for months and then move quickly once the thing underneath them is built. Weight, pipeline, and revenue all behave differently, and the difference is knowable before you start.
If you know the shape, a flat month reads for what it is. Either it is what this measure does at this stage, or it is a departure from what you expected, and you are not negotiating the answer in the moment.
What it costs to have neither
I have let plans run for six to nine months without honestly asking whether they were working. Not through neglect. Every individual month had a defensible reason to hold, and the reasons were all true. What was missing was any prior view of when the number should have started to move, so there was never a month where holding became indefensible.
Now it is rarely longer than a month. The difference is not discipline. It is that the question gets asked on a schedule rather than when something feels wrong, and feeling wrong is a terrible trigger. It arrives late, and it arrives on the plans you were least attached to rather than the ones going least well.
A review does not tell you whether to change the plan. It tells you whether you have earned the right to decide yet. Early on, usually you have not, and the useful output of the review is not a decision but a sharper idea of when the number should begin to move. The plans that waste a year are not the ones changed too slowly. They are the ones where nobody ever established what moving would have looked like.