Sending an application and delivering one are different events, and almost no tool measures the second
Every tool in this category reports the same thing back to you: it filled the fields, it attached the file, it clicked the button. All three of those are the tool reading back its own input. None of them is the employer saying anything.
We shipped the boring version of the fix. After the submit, we wait for the employer's system to answer, and we show three states instead of one: landed, still verifying, did not land. The third column is the one that makes the numbers look worse and the product worth using.
Launched on TinyLaunch this morning.
The open question I genuinely cannot answer: is a smaller honest number actually more persuasive to a buyer than a bigger vague one, or does it just read as a weaker product?
Smaller honest number only reads weak when it floats without context. A "did not land" column has built-in context — it says you measure outcomes, not your own clicks. We published a €150-to-zero-installs breakdown once; the honest number got more real replies than any metric post we've done. At this stage trust is the feature.
Honestly, yes — it pulled more from strangers than from the handful who already knew the story. People in orbit had heard it once; for everyone else the post was news. That's the testable bit you want: honest numbers are news, vague ones are noise. A cold reader sees "did not land" and learns something about the product; nothing to distrust yet.
"Honest numbers are news, vague ones are noise" is doing a lot of work in one line, and the cold-versus-warm split is the part I did not expect to be so clean. It makes sense in hindsight. People already in orbit have spent their trust in advance, so the evidence is redundant to them. A stranger has spent nothing yet, which is the only state where evidence can actually move something.
The corollary I am now sitting with is that this has a shelf life. The first honest number is news. The second one is a baseline, and from then on you are not being read for candour, you are being read for the direction of travel. Which is fine, and it is a much better position than the one before it, but it is a different job. Publish once and it is a story. Publish continuously and you have signed up to a scoreboard that can go the wrong way in public.
I think that is the actual cost of trust-as-a-feature, and nobody warns you about it because the first post feels so good. You have gone further down that road than I have. Did the number staying public change what you were willing to try, or did it turn out to matter less than it sounds like it would?
That reframe is better than mine, thank you. "Floats without context" is exactly the failure mode. A bare "did not land" is a confession. The same number sitting next to a reason is a status.
We learned that the slow way, because the first version of the column just said failed, and it read as us apologising for the product. Now each row carries the stage it stopped at and a reason code, and whether the next move is ours or genuinely blocked at the employer's end. Identical number, completely different feeling to read.
"At this stage trust is the feature" is going to stick with me. It also predicts something testable: if trust is the feature, the honest number should convert better on a cold audience than a warm one, because people who already follow you extend the trust for free and don't need the evidence. No idea whether that actually holds. Did your 150 euro breakdown pull more from strangers than from people already in your orbit?