Acquisitions company
Custom acquisitions CRM with multichannel campaigns
A CRM that turns public county records into scored seller leads, runs automated mail, SMS and email campaigns, and triages every reply with AI.
- parcels ingested from county records across three counties
- 490K+
- parcels ingested from county records across three counties
- owners scored as prospects
- 127K
- owners scored as prospects
- active contacts in automated campaigns
- 22K+
- active contacts in automated campaigns
- letters, postcards, SMS and email in one sequence per lead
- 4 channels
- letters, postcards, SMS and email in one sequence per lead
The challenge
Off-market acquisitions depend on reaching the right owners before anyone else, repeatedly and across channels.
Off-the-shelf CRMs could not ingest county data, score owners the way we underwrite, or coordinate mail and digital touches in one sequence. Replies flooded in, and most were noise.
What we built
01
Ingest and score
Public county records are ingested, cleaned and matched into one owner and property record, then scored on how likely the owner is to sell and how well the home fits.
02
Automated drip campaigns
Each lead enters a multichannel sequence of letters, postcards, SMS and email, timed and personalized automatically. Do-not-contact and litigation-risk numbers are excluded, with cooldowns and staggered sends.
03
AI triage for every reply
A language model reads each inbound reply with the deal context, classifies it, drafts a response and suggests a callback time. Opt-outs and dead ends close instantly; interested sellers go to a person with the context summarized.
Prospect score breakdown · 90 pts scored Jun 1
| Signal | Points | Detail |
|---|---|---|
| Tax delinquency | 0 | — |
| Deed distress | 0 | — |
| Absentee owner | 20 | Mailing address differs from property |
| Ownership tenure | 30 | 28 years since last sale |
| Equity estimate | 0 | Mid tier, no lien signal |
| Value trend | 0 | Stable |
| Building age | 5 | Built 1960 |
| Neighborhood value | 35 | $/sqft in the bottom 10% of nearby homes |
Timeline
Skip traced: 2 phones (1 DNC), 1 email
system · Jul 7
Suppression updated: ⊘ Mail · ✓ Calls · ⊘ Email
J. Rivera · Jun 23
Interested: New → Interested
J. Rivera · Jun 23
Direct mail delivered (Round 1)
system · Jun 23
Direct mail sent (Round 1)
system · Jun 6
Deal created from parcel
system · May 27
What we learned
Automation makes a channel cheaper to test, not automatically profitable. Measuring it honestly told us where to invest and where to stop.
Stack
- TypeScript
- PostgreSQL
- Cloud Tasks
- LLMs
- Telephony and direct-mail APIs
What would you build with one more great operator on the team?
A 30-minute call. We'll ask where the time goes and tell you honestly whether AI can help.