Eleven thousand emails. Nine replies. Zero customers.
Proofdrop exists because I ran the numbers on my own failure and found one uncomfortable pattern in it. This page is the long version, including the list of things the product refuses to do, and that list is the spine of everything else.
Why this exists
Two years ago I was selling a logging tool and doing everything the playbooks say. Good list, personalised first lines, four-step sequences, warmed domains. Eleven thousand emails produced nine replies and zero customers. I stopped because the math was embarrassing, not because I ran out of list.
When I finally sat down with the data, the nine replies had one thing in common. They were the emails where I had done something first. Run their signup flow, found a broken redirect, timed their API, and led with what I found. Those took forty minutes each to write, which is why I only ever sent a handful.
The other ten thousand nine hundred and ninety-one emails had a different thing in common: they were addressed to people who did not have the problem I fixed. No subject line survives that. The failure was never my copy. It was that the email carried no proof I existed, sent to someone who had no reason to care, at a moment when they were not listening.
So the question became mechanical: what would it take to do the forty-minute version every morning, for every product that launched that day, without a human in the loop? Proofdrop is the answer. It watches the launch directories, matches what launched against what you sell, and asks your own system to produce one real, checkable finding per company. Your tool, your probes, your voice. We run the boring machinery: sending, pacing, caps, de-duplication, bounces, unsubscribes. That decides whether a good email actually arrives.
What Proofdrop refuses to do
These are not guidelines. They are enforced in the pipeline, and an account that asks for them gets a no, not a workaround.
No sending without evidence
Every email must carry one specific, timestamped, checkable finding about the recipient’s own product, produced by the customer’s own system. If a run finds nothing worth saying, nothing is sent. The pipeline does not fall back to a pitch.
No scraped or bought lists
Recipients come from public launches we swept that morning (directories, Show HN, Product Hunt) and matched against what the customer sells. There is no upload-a-CSV feature. There never will be.
No invented evidence
A finding is a real probe result with request ids the recipient can verify. We do not generate plausible-looking numbers, and we do not let customers paste in claims that did not come from a run. Fake proof is worse than no proof.
No unlimited volume
Daily caps hold at 40 sends per account on a busy day, with spacing between sends. The cap exists because the whole point is that each email was worth sending. Volume is the thing we are built against.
Who is behind this
Proofdrop is built and run by Carol Campistol. The support inbox is answered by a person, usually the same one. If your cold email is producing nothing, the problem is probably not your copy. It is that your email carries no proof you exist. Fix that and the first week of replies looks strange: founders thanking you, asking questions, forwarding it to their co-founder. That is what this is for.
Questions: support@mail.proofdrop.net. The full sending rules live on the sending policy page.