The message test (Track 1)

The goal is not to close anyone by email. It is to learn which description of the problem earns a reply from a university lab. You run three versions of the same offer and count which one brings back the most holograms.

The rules that do not change between the three arms

• Same offer, same ask. The ask is small and free: send us one hologram and we will reconstruct it and send the result back. No call, no commitment, no account.
• Individual plain-text emails from a real mailbox. No HTML template, no attachment, no tracking pixel — a pixel is treated like a cookie in some jurisdictions and, more importantly, labs notice and distrust it. (Current outreach guidance agrees: first touches should look low-risk — plain text, no links where avoidable.)
• About 20 a day. Not a blast. Personal, one at a time.
• The metric is holograms received per arm — not opens (you are not tracking), not even replies. A hologram arriving is the signal.
• One variable only. The three arms differ only in how they name the pain. Keep subject style, length, and the ask identical, or you learn nothing.

The one-owner-per-row rule (do this before you send anything)

Every prospect row has exactly one owner. Before you contact anyone, open the prospect log (page 12) and put your name and the date in the owner column of each row you will work. If a row already has an owner, it is not yours — leave it. This is how two people never email the same researcher. A send with no claimed row is a process error; fix the log first.

The three arms (each names a different pain)

Every word below is inside the claims boundary. No accuracy claim, no agreement number, no speed number appears — because none is needed to offer a free reconstruction.

Arm A — continuity.

Subject: reconstructing your holograms without the person who set it up

Hi Dr. ⟨name⟩ — I run a hosted service that turns off-axis holograms into height maps and a provenance record (a receipt of exactly which software and settings produced the result). Labs tell us the setup usually lives in one student's head. If you send me one hologram with its wavelength, pixel pitch and rough distance, I'll reconstruct it and send back the result and the receipt — no account, no call. Worth a try?

Arm B — provenance.

Subject: a re-computable receipt for your reconstruction

Hi Dr. ⟨name⟩ — when a reviewer asks exactly how a height map was produced, can you hand them something that re-computes to the same answer? Our service returns a content-addressed provenance record with every reconstruction: it re-hashes, names the software version, and commits to your input and settings without us publishing them. Send me one hologram (wavelength, pixel pitch, distance) and I'll send back the result and that record. Free, no account.

Arm C — reproducibility.

Subject: same hologram, same answer, every time

Hi Dr. ⟨name⟩ — we host a reconstruction service built so the same input and the same config always produce the same result ID. If getting the identical number twice has ever been a problem, send me one hologram with its wavelength, pixel pitch and distance and I'll reconstruct it and send the result back with its provenance record. No account, no call.

What to log

In the prospect row: arm sent (A/B/C), date, and whether a hologram came back. Roll the per-arm counts into the weekly report (page 09). Success = holograms arriving and one arm clearly ahead. Give it 2–3 weeks. When a hologram arrives, you are on page 17 within 48 hours.

Artifact when done: a per-arm holograms-received count in the prospect log, and every contacted row showing a single owner and a sent date.

Buyer pages to pair with this page: workflow fit · biomedical QPI · holography labs · industrial QA. Pipeline stages and KPIs: page 00.