Dispo Deal Blast

Send one deal to the buyers who actually buy at that price, then turn what comes back into an asset you keep.

Most dispo lists are a spreadsheet of everyone who ever bought a house. This builds the opposite: a small list of people whose recorded purchases say they buy this kind of property at this kind of price, reached on a number they have heard from before.

Use when someone says "blast this deal", "who do I send this to", "text my buyers", or "set up dispo texting".

Requirements

• Python 3.10+ (the bundled script is stdlib only)
• A CRM or data source with deed-level investor transactions
• An SMS provider with numbers you own, and a compliance posture you understand

Everything below is the method. None of it needs a key to read, and the bundled calculator runs on plain Python.

The chain

registry  ->  recency gate  ->  price band  ->  copy  ->  staged blast  ->  phonebook

1. Registry. Sweep investor transactions in your county. Hydrate each property for the deed owner, because search results usually carry no name. Dedupe by mailing address. Resolve LLC principals.
2. Recency gate. Re-run the sweep bounded to the last 12 months. Recency decides WHO is on the list; full history decides WHAT they buy.
3. Price band. Each buyer's real purchase range, p10 to p90, from priced sales only.
4. Copy. One message per buyer, audited before anything is staged.
5. Staged blast. Everything lands as HELD. Release is a separate human act.
6. Phonebook. Classify every reply, suppress the outs, keep the buyers.

The rules that matter

Each of these cost a real failure to learn. They are the actual content of this skill; the code around them is easy.

The semantic trap that ruins the list

Transaction labels describe the LAST SALE, not the current owner.

• pending, wholesale, wholetail, rental mean the investor still holds, so the current owner IS the buyer. These are your targets.
• flip means the exit already happened, so the current owner is a retail homebuyer. The investor you want is the last-sale SELLER.

Texting the flip bucket reaches ordinary families who just bought a house.

Probe every filter by count delta

An unknown filter key is silently ignored. A deliberately bogus key returned a byte-identical count to no filter at all. Acceptance proves nothing: change one filter, compare the count, and only then trust it.

The same applies to search scoping. A county field that wants "Knox" returns zero for "Knox County, TN", and zero is indistinguishable from an empty segment.

The price band is asymmetric

Above a buyer's band is affordability: someone whose ceiling is $120k cannot close $600k. Below it is only interest: someone whose cheapest purchase was $140k can obviously afford $75k, they may just not want one that small.

A symmetric tolerance kept 68 of 199 buyers at a $75,000 ask and dropped people for being too big, which is the wrong reason to skip anyone. Divide the floor by a multiple; keep the tolerance on the ceiling.

"No institutional buyers" cannot be a keyword rule

Scanning for HOMES / BUILDERS / CONSTRUCTION flagged 24 names on a real cohort. 23 were small local operators with 1 to 11 doors. A self-performing local builder is the single best buyer for a heavy rehab, so the keyword sweep deletes the target audience to catch one name. The same failure hits BANK inside WILLBANKS.

Use a named list of firms you can verify are production builders, iBuyers or SFR funds, and log every drop with the name it matched.

Address discipline, both directions

Default to withholding the address: name the road, and let the OFFER do the redacting ("I can send the walkthrough, photos and full address if you're interested"). Saying it defensively ("not posting the address here") draws attention to the withholding.

Disclosing the address is an explicit per-deal opt-in, never a code change, so the redaction default keeps protecting every other campaign.

The price whitelist beats a blocklist

Never state a figure other than the approved asking price. Implement it as a whitelist: extract every money-looking token and require each to equal the approved price. A blocklist only catches the number you remembered to ban; a whitelist catches the contract price without the audit ever being told what it is, which means the contract price never has to enter the process.

Copy is audited on every message, not a sample

Variants are chosen by a hash of the record, so the four messages a sample prints are not the four that ship. Three variants once shipped unsigned and only two surfaced in a sample. Assert on all of them, and refuse to stage on failure rather than printing a warning above a staged batch.

Sticky senders

A buyer who replied to one number and then hears from another reads as a spam farm, and a callback rings a line with no history of them. Pin each buyer to the number that actually texted them last, and verify it by joining the new batch against the old one before releasing.

Staging twice is the worst thing a cold number can do

Staging is a WRITE. A dry-run flag that gates sends but not staging will report itself as a dry run while writing the whole batch. Add a duplicate guard that skips any phone already holding a pending message, and return the count so a caller cannot report more than it staged.

Human gates

Keep these with a person, permanently:

• Release. Staging sends nothing; release is the irreversible step.
• The address, photos and lockbox code. The agent offers them; a human sends them. A lockbox code is physical access to a house and never belongs in an agent's facts.
• Any reply that names a price. Negotiation is not an agent's job.

Files

• references/buyer-registry.md — sourcing, dedupe and principal resolution
• references/sms-guards.md — every gate and the failure it prevents
• references/copy-rules.md — the voice rules and the audit
• scripts/cohort.py — stdlib cohort and band calculator