Observed from public sources. Normalized for consistency. No rankings. No promotion. No lead capture.
Read methodology →
| Observed | Title | Type | State | Price | Source |
|---|---|---|---|---|---|
| Loading listings... | |||||
| Observed | Title | Type | State | Price | Source |
|---|---|---|---|---|---|
| Loading listings... | |||||
Know a business broker with listings online? Submit their URL. Our AI agent analyzes the page, generates a scraper, and adds them to the ledger.
Automated scrapers visit 1,687 broker websites daily. Specialized parsers for major franchises. ML pattern detection for the long tail.
Every listing gets a hash, a timestamp, and a source link. If it can not be traced to its original broker, it does not exist in the ledger.
Snapshots are committed daily. Diffs show what is new, removed, and relisted. The full history is public and reproducible from code.
Every snapshot includes SHA-256 hashes. Anyone can rerun the scrapers and compare. The code and methodology are open.
→ Source-linked. Every record traces back to its original broker page.
→ Timestamped. First seen, last seen, and every change in between.
→ Reproducible. The code is open. The methodology is documented. Run it yourself.
→ Append-only. We do not edit history. Removals are recorded, not deleted.
→ No opinions. We do not rank brokers. We publish what we observe.
DealLedger is open source. Fork it. Extend it. Add a vertical. Challenge the methodology.
The problem: There is no public registry for business listings. The same listing can appear multiple times across sites, disappear without notice, and persist after removal. Inventory is not auditable.
What this is: DealLedger is a public deed registry for businesses for sale. We scrape broker websites, timestamp everything, and publish the raw data. Every record is source-linked. Some records are broker-verified; others are marketplace-only and labeled as such.
What this isn't: We don't rank listings. We don't recommend brokers. We don't capture leads. We don't broker deals. This is a record, not a marketplace.
How it works: Automated scrapers pull listings from broker websites. Data is normalized, deduplicated, and published as flat files. The code is open source. The methodology is documented. Anyone can verify. Coverage expands over time.