The problem this solves
A meaningful share of product signups arrive on personal email domains, and every one of them is a company hiding behind a webmail address. The person evaluating your product on a Tuesday night is often doing it for their employer, but the CRM sees a consumer address with no company, no size, no segment, and routes it to nobody. These are frequently the warmest signals a pipeline gets, and they sit unworked precisely because the one field that would make them workable is missing.
Resolving them by hand is detective work: search the name, cross-reference the profile, guess among the five people who share it, and hope. Done properly it takes minutes per signup, which nobody has at volume, so teams either ignore the segment entirely or import it raw and let the junk rows pile up. Either way the CRM ends up misrepresenting the actual inbound interest the product is generating.
The workable version is a mission that treats resolution as an evidence problem: identify the person behind the address with real corroboration, resolve their employer, and write the result into the CRM only when the match clears a confidence bar, with the evidence attached. Low-confidence guesses stay out of the database, because a wrong company on a hot signup misroutes the exact lead you most wanted to get right.
How the mission runs
- Collect the personal-domain signups. The mission starts from your signup list, filtered to personal email domains. Each row carries whatever the signup captured - name, address, sometimes little else - and that sparse starting point is exactly what the rest of the pipeline exists to fix.
- Resolve the person behind the address. Enrichment works each signup to establish who the person actually is: matching name and address against professional footprints until a real identity emerges with corroborating evidence. Signups that resolve to multiple plausible people are held at this stage rather than forced to a coin-flip conclusion.
- Resolve the employer. For each resolved person, the same evidence trail establishes where they work now: company, role, and seniority. Current employment is the fact that matters, since a signup from someone's previous-job identity routes to the wrong account entirely. Stale matches are flagged instead of written.
- Write people and companies into Attio. Resolved people and their companies are written into Attio, deduplicated against what already exists so a returning signup enriches an existing record instead of spawning a twin. Writes are idempotent, and every created or changed record notes the evidence that produced the match.
- Report the resolution ledger. The mission closes with a ledger: how many signups resolved cleanly, how many matched existing Attio records, and the held-out remainder with the reason each stalled - ambiguous identity, stale employment, or insufficient evidence. The unresolved set stays visible for a retry or a human look.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
An Attio workspace where personal-domain signups have become real records: people with names, roles, and current employers, linked to company records, deduplicated against everything that already existed, and each carrying a note on the evidence behind the match. Plus a resolution ledger showing what resolved, what merged into existing records, and what was held out with reasons, so the residue is a queue rather than a mystery.
Make it yours
- Run it as a standing daily sweep on new signups so resolution happens within a day of the signup, while the person still remembers signing up.
- Add a routing step: resolved companies matching your target segments get flagged for sales in the ledger, turning the resolution pass into a lead triage pass as well.
- Point it backward at the historical signup list first; the backlog usually contains a surprising number of now-obvious accounts that signed up before anyone was looking.
Frequently asked questions
What stops a wrong match from landing in Attio?
A confidence bar. A match is written only when the evidence corroborates one identity and one current employer; anything ambiguous is held out and listed in the ledger with its candidates. The mission treats a blank as recoverable and a wrong match as expensive, and it errs accordingly.
Will this create duplicates of people already in Attio?
No. Every write is deduplicated against existing Attio people and companies first, so a returning signup or an already-known contact gets enriched in place. Writes are idempotent, which means re-running the mission over the same signup list never multiplies records.
Is resolving people from a signup appropriate?
The mission works from information people have made professionally public, connecting a signup someone chose to make with the professional identity they maintain. It records evidence rather than speculation, holds out uncertain cases, and writes into your CRM under your existing data-handling policies, which you should apply as you would to any enrichment.