The data behind customer acquisition on every channel: lookalikes of your best customers for outbound and paid, retargeting and winback audiences from the CRM, suppression so budget reaches new accounts only, and one scored list feeding sequences and ad platforms alike. The infrastructure under acquisition, run by autonomous AI agents.
Lookalikes of customers, retargeting of open and stalled deals, winback of churned and closed-lost, each built from live records.
The same scored, suppressed list loads a paused sequence and a matched audience, so outbound and paid agree on the accounts.
Customers and open opportunities suppressed before any audience or sequence is built.
Every load into a sequence tool, CRM or ad account waits for a person; research never does.
Sales built a list from one tool and marketing built an audience from another. The account that receives a first touch on Tuesday never sees the ad, and the one that sees the ad has no one calling.
Closed-won records describe the ideal customer better than any persona doc, closed-lost records describe the winback list, and neither is queried because turning records into audiences takes an export and an afternoon.
Customers see the acquisition ad, open opportunities get the cold sequence, and churned accounts are treated like strangers. The suppression list exists in a spreadsheet that nobody uploaded this month.
Closed-won accounts read from the CRM and expanded to the companies nearest them on industry, size band, stack, funding and signals, scored against the ICP with the resemblance explained.
Open and stalled opportunities, closed-lost from the last year and churned customers pulled as their own lists, re-enriched so the contact is still the right person and the company is still the same size.
Customers, open opportunities, competitors and the workspace suppression list removed from every acquisition list before it leaves, and pushed as exclusion audiences where the platform supports them.
The same scored rows loaded into Lemlist or Instantly as a paused sequence, into LinkedIn Ads and Meta Ads as matched audiences, and into the CRM as a campaign membership, after approval.
Customers, open opportunities, closed-lost and churned pulled from HubSpot, Pipedrive, Attio or Zoho CRM with their owners and dates.
Lookalikes expanded from customers; retargeting and winback lists cut from opportunities and churn with the recency you set.
Every company re-enriched and every contact re-verified, so the winback goes to who is there now and the lookalike has a named buyer.
Fit against the ICP with reasons, and for winback and retargeting the signal since the last touch: new funding, new leadership, new hiring.
Customers, open deals, competitors and the suppression list removed; exclusion audiences prepared for the ad platforms.
Sequences paused in your outreach tool, matched audiences in LinkedIn and Meta, a Customer Match export for the rest, each after one approval.
Two hundred lookalikes of the last fifty customers, the buying committee verified at each, loaded into Instantly paused and pushed to LinkedIn as a company audience the same morning.
Closed-lost from the last twelve months re-enriched: who is still there, what changed, a new round or a new leader, and a draft that references the change, staged for the rep who owned the deal.
Opportunities with no activity in 30 days pushed to Meta as a custom audience of the actual decision-makers, so the deal stays visible while the rep works the next step.
Customers, open deals and the do-not-contact list synced as exclusion audiences to LinkedIn and Meta every week, and applied to every sequence load.
| By hand | With AstroFabric | |
|---|---|---|
| Source of truth | Two lists from two tools | One scored list from the CRM and licensed data |
| Lookalikes | A persona doc | Nearest neighbors of closed-won, explained |
| Winback | Closed-lost left alone | Re-enriched, with the change since the deal |
| Suppression | A spreadsheet uploaded sometimes | Applied on every build, synced as exclusions |
| Channels | Outbound and paid on different accounts | Sequence and audience from the same rows |
| Refresh | Quarterly, by hand | Weekly, after one approval |
Customers expanded to two hundred lookalikes, the committee verified, loaded into Lemlist paused and pushed to LinkedIn as a company audience after one approval.
Closed-lost from the last year re-enriched, the signal since the deal attached, drafts staged per rep and a Meta audience of the same people.
Customers, open opportunities and the do-not-contact list pushed as exclusion audiences to LinkedIn and Meta every Monday, with the diff reported to Slack.
Three of a hundred. The full playbook library ships in every workspace, and anything you can describe becomes a mission.
Reading the CRM, building, enriching and scoring run inside the credit ceiling. A push into a sequence tool, an ad account or the CRM is staged for approval as previewed.
Customers, open opportunities, competitors and the workspace suppression list are removed before any acquisition list is delivered, and the removed count is shown at the gate.
Sequences load paused and audiences are created; your team sends the messages and runs the campaigns inside the tools they already use.
Lookalike and contact data come from licensed providers. The agents never scrape member networks.
The data layer that feeds every acquisition channel from one place: lookalikes of your customers, retargeting and winback lists cut from the CRM, all re-enriched, verified, scored and suppressed, then loaded into sequences and ad platforms after approval so outbound and paid work the same accounts.
The CRM provides customers, opportunities, closed-lost and churn with their dates and owners. Licensed sources provide the lookalike expansion, fresh firmographics and technographics, the signals since the last touch and verified contacts. The agents join the two on canonical domain.
Closed-lost and churned accounts within the recency you choose are re-enriched, the contacts re-verified and replaced where people moved, and the changes since the deal attached: funding, leadership, hiring, stack. Rows with a real change rank first, and drafts can be staged per owner.
Lemlist and Instantly as paused sequences, LinkedIn Ads and Meta Ads as matched or exclusion audiences, HubSpot, Pipedrive, Attio and Zoho CRM as campaign membership, a Google Customer Match export for upload, and CSV or Google Sheets, each after one approval.
No. Sequences load paused in Lemlist or Instantly and audiences are created in LinkedIn Ads and Meta Ads; sending, creative, budgets and campaign settings stay with your team inside those tools. The agents prepare every input and a person takes every action that reaches a buyer.
Yes. A schedule re-reads the CRM, rebuilds lookalike, retargeting, winback and suppression lists, re-verifies contacts and stages the pushes. Each week’s approval takes a minute and the diff since last week is reported to Slack. Pause or edit the schedule in plain language at any time.
Lookalike discovery and enrichment are a few credits per company, a verified contact is a few credits, and CRM reads, scoring, suppression and pushes are free. Rates show before each run, misses on contact finding cost nothing, and plans start at $49 per month.
Connect the CRM. Lookalike, retargeting and winback lists come back scored and suppressed, ready for every channel after one approval.