Research every target before you reach them.
Dossier-First Outreach pulls your leads from the CRM, builds a verified, enriched dossier for each one before a single line is drafted, and then sends personalized outreach that is grounded in that research. Nothing leaves the building until it clears two gates, on-brand and, where your cadence says so, signed off by a human.
Run it on your own batch of leads See a dossier and a gate log
THE PROBLEM
Most outreach starts from a template and mail-merges a first name into it, which everyone can tell, so it gets ignored (or it gets a reply you did not want). The newer answer, an AI SDR that writes and sends on its own, fixes the personalization but removes the one thing you actually needed, a human between the model and the send button. So you are stuck choosing between cold templates that waste the list and autonomous senders that can put an off-brand or off-key message in front of exactly the person you were trying to win. Neither one does the boring thing that works, which is researching the target first and then having someone check the message before it goes.
HOW IT WORKS
It runs in one order, and the order is the point. First it pulls the batch from the CRM source you configure. Then, for each lead, it builds the dossier (enrichment, a real target profile, shepherd assignment) before anything is written, so every message is grounded in research rather than a merge field. Then it drafts the outreach against that dossier and runs it through two fail-closed gates, a brand-voice check and a human-in-the-loop requirement check. Only a message that clears both gates sends. Anything that fails the brand check, or that your cadence says needs a human and does not have one, is held, not sent, and you see why.
WHAT YOU GET
For each lead in the batch: - a verified dossier, built first, that grounds the outreach in a real profile instead of a template, - a personalized message that references the same narrative as the dossier, - a gate decision, on-brand or held (with the reason), and whether a human sign-off was required and present, - a send log you can read after the fact, what went out, what was held, and why.
Volume that never costs you the brand, or the human in the loop.
WHO IT'S FOR
Founder-led and small go-to-market teams running outreach off a CRM who need it personalized and on-brand at volume without a full SDR bench, first, then agencies and RevOps teams managing outreach for several brands where one off-voice message is a real problem. If you care that the message is researched, on-brand, and checked before it sends, this is for you. If you are running a spray-and-pray blast that does not care about any of that, it is not.
PRICING
Prices below are hypotheses we are validating, not commitments, and we price the ladder, not a single number. - Pilot, a fixed-price paid run on your own batch, where you see the dossiers, the drafts, and every gate decision before you commit to anything. - Metered, per researched-and-reached lead, once the pilot proves the personalization and the gate discipline. - Platform, a seat or team license once you are running the motion continuously across segments, with brand voice and the human gate configured per brand.
Every outreach send is an outward action, so it is fail-closed, a human sign-off is required before any message leaves when your cadence demands it.
PROOF (the dogfood)
We run this on ourselves. AYA's own go-to-market pipeline, design partners and regulated-AI buyers, is pulled from our CRM, each lead is enriched into a real dossier before we write anything, and every message we send passes the same brand-voice gate and human-in-the-loop gate we are describing here. The AYA company is the first customer of its own outreach line.
HONEST NOTE
This is a structural build, and we will tell you exactly what that means. The pattern is built and verified, its pieces are the same Kendra outreach handlers we already use, and it composes them dossier-first with the two gates in place. What has not happened yet is a live run, we have not pulled a real CRM, enriched a real lead, or sent a real email through it end to end, because that needs live credentials and a running system, and this bundle deliberately does no live sending and no spend. So the honest status is built and gated, not yet run against a live list. When we say the gates are fail-closed, that is a design property you can read in the pattern, and the first live batch is the natural next step once the credentials are in place.
*This page is a specification. The capability it describes is not built yet, and nothing here is a claim that it runs today.*