Two fixed points, the same for every campaign, and the reasoning behind them so you can check it.

A letter only counts once it has had two weeks to get an answer. Before that, silence means nothing: the letter may still be sitting in a pile on a desk. So every number on this page counts letters that have had their two weeks, never letters just posted.

Each letter is followed by up to 3 nudges on LinkedIn or by email. The bar a working campaign has to clear is 5% of people writing back. That is a bar we set, not an average we measured, and not a promise about what yours will do.

The two points where you get a result

Letters	What you learn
48	A first feeling for who reads. You see how many people scanned the letter, which roles and companies they come from, and who wrote back, as plain counts such as “3 of 60 answered”. No verdict yet, either way: on this few letters a quiet stretch and a weak letter look the same.
150	The verdict, one of three. It works: enough people write back. It does not work: so few answer that bad luck no longer explains it. Or it is unclear: fewer answers than a working campaign gets, but not few enough to rule the letter out. Then we change one thing and start counting again.

Why the numbers are fixed

Every campaign gets the same two numbers. A finish line that moved with each campaign felt like a goalpost being moved, so we fixed it. You always see how far along you are, as “20 of 48” or “90 of 150”, never a guess at how many more it might take.

The follow-ups are why 150 letters are enough. The same person ignoring you four times is one answer, not four. What the follow-ups buy is more people answering, and when answers are rare, more answers per letter does the same work as more letters. A letter sent on its own would need far more of them to tell you the same thing.

What this cannot do

It cannot pit two letters against each other and name a winner. Telling two similar results apart takes far more letters than a campaign sends in months, so any ranking of letter A against letter B would be noise with a percentage sign on it. We change angles on judgement and on what people say back, and we tell you that is what we are doing rather than dressing it up as a test.

It also cannot squeeze a verdict out of a small number by looking at it harder. Before 48 letters have had their two weeks there is nothing to read yet, and between 48 and 150 the honest answer to “does it work” is still: too early. That is what the weekly read will say, in those words.

The mistake we accept

We accept about a 5% chance of calling a working letter a failure. Asking for more certainty would mean waiting much longer for every verdict. That is the point we picked, and you are entitled to hold us to it.

You will feel one thing in the first weeks. Good news shows up early and bad news takes longer, because you are counting something that rarely happens either way. A letter that works well stands out within the first few dozen. Silence needs many more letters before it says anything with the same confidence. That is why the first read never calls a letter dead.

In weeks, on the plan with 100 people a month: at 100 letters a month, the first read comes after about 5 weeks and the verdict after about 9, because the last letters counted still need their two weeks. The verdict is where you decide: keep going, change one thing, or stop.

The other thing worth checking

The letters are printed, not written with a pen. An algorithm varies every stroke, so no two pages match, and I will not claim more than that. We don’t put it on the envelope, because the handwriting has one job, to get the letter opened. Everything after that can be said out loud, and when a prospect asks, the true answer is also the strong one: the handwriting was machine-made, the decision to write to them was not.

If this does not survive your arithmetic

Tell me and I will fix the page. An earlier version promised an answer sooner than the letters could carry one, and every campaign had its own finish line. Both are gone: two fixed points, the same for everyone. That correction is the reason this page exists.