A baseline built from four posts is a rumour. A baseline built from forty is measuring an account that may have changed halfway through. Somewhere between those is a usable window, and the honest answer is that it depends on how much the account varies.
The thing that actually sets the threshold
Not the post count. The spread. An account whose posts land within a narrow band reaches a stable baseline quickly, because each additional post confirms what the previous ones said. An account whose posts swing wildly needs many more before the middle stops moving.
This is why a single universal number is always slightly wrong. The question is not “have I got enough posts” but “has the baseline stopped moving as I add them”.
A practical test you can run by hand
Compute the baseline over your last n posts. Then compute it again over your last n minus five. If the two numbers are close, the baseline is stable and you can trust multiples against it. If they are far apart, it is not stable yet, and any multiple you calculate is mostly noise.
That test costs nothing and answers the question directly for your account rather than for accounts in general. It is also the test we would want anyone to apply to our numbers.
The baseline is ready when adding posts stops changing it. Not when it reaches a particular count.
Rough expectations
For most accounts posting consistently in one format, a window in the region of fifteen to thirty recent posts is where the baseline tends to settle enough to be worth using. Below roughly ten, treat any multiple as directional at best — it will tell you that something unusual happened, not how unusual.
These are working expectations, not laws. The test above beats the range every time, and where they disagree, believe the test.
What to do when you are below the threshold
Three options, in order of usefulness.
- Use rank rather than ratio. “This was my best post of the last twelve” is a weaker claim than a multiple, but it is a claim your data can actually support.
- Widen the window and accept the lag. More history buys stability at the cost of currency. For a new account that has not changed much, this is a good trade.
- Wait. Genuinely the right answer more often than people want. A baseline is cheap to acquire — it accumulates on its own while you post.
The failure this prevents
Without a stability check, an account with six posts can produce a post at “8× baseline” that means nothing at all, because the baseline was set by five quiet posts and one of them was the account's first. That number will then get used to justify a format change.
This is the most common way measurement makes decisions worse rather than better: it produces a confident figure from data that could not support one. For what the figure means once it is trustworthy, see how to read a multiple.