LearnDealer and channel data

Duplicate business entities across dealer networks

By Published 3 min read

Explore with AI

The short answer

Duplicate business entities appear when one physical location exists as several records across manufacturer, distributor and map systems, each with its own identifier. An assistant treats them as separate businesses and may combine their facts or choose the wrong one. The remedy is a canonical record that every other identifier points to.

In this article · 6
  1. How one location becomes several
  2. What a machine does with duplicates
  3. Entity resolution is a decision, not a match
  4. The canonical record
  5. Where to look first
  6. Questions

How one location becomes several

Duplicates are rarely a data-entry mistake. They are the natural result of systems that each mint their own identifier for the same place.

  • A change of ownership or a rebrand creates a new dealer record while the old one stays active in feeds.
  • A relocation creates a record at the new address without retiring the old one.
  • A parts counter and a showroom are entered as two locations in one system and one in another.
  • A distributor account number, a manufacturer dealer code and a map listing identifier are never mapped to each other.
  • An acquisition merges two companies' systems without merging their records.

What a machine does with duplicates

Each record carries its own partial facts. One has the right phone and an old address. Another has the new address and no hours. A third has the services but the former brand. A system that cannot tell they describe one place will present them as alternatives, merge fields across them, or favor the one with the most inbound references, which is usually the oldest.

The failure is not a wrong value. It is a wrong count and a wrong history: the assistant lists two stores where there is one, and attaches the old store's lapsed authorization to the new store's address.

Entity resolution is a decision, not a match

Matching names and addresses finds candidates. Deciding that two records are the same place is a judgment with consequences, because merging deletes the independent history of one of them. The judgment needs evidence: a shared legal entity, a continuous phone number, a documented relocation, a date.

The result should be recorded, not just executed. A canonical record states which identifiers it absorbs, why, on whose confirmation and when. Without that, a later feed re-creates the duplicate and nobody can say why it was merged.

The canonical record

The durable fix is a single canonical record per physical location and per legal entity, with a mapping table from every external identifier (manufacturer dealer code, distributor account, map listing, aggregator ids) to that record. Every outbound feed publishes the canonical values. Every inbound copy is reconciled against it.

A retired identifier is not deleted. It is marked retired, with the date and the canonical record that replaced it, so a feed that still carries it can be recognized as stale.

Where to look first

Start with locations that changed: acquisitions, relocations, rebrands and lines added or dropped in the last several years. That is where the duplicates are. Compare each against the identifiers present in the manufacturer locator, the distributor portal and the main map listing, and list the ones that map to more than one record.

Questions

Why do duplicate records matter to AI assistants?

An assistant that cannot tell two records describe one place may list two stores, merge their fields or attach an old authorization to a new address.

Should duplicate records simply be deleted?

No. Retire them with a date and a pointer to the canonical record, so a feed that still carries the old identifier is recognized as stale.

What evidence supports merging two records?

A shared legal entity, a continuous phone number, a documented relocation or ownership change, each with a date and the person who confirmed it.

Is this only a problem for large networks?

It scales with the number of systems that mint identifiers. A single dealer with a manufacturer code, a distributor account and a map listing already has three.

Explore with AI

Where does your own information disagree?

Get Upstream

One email a week: what we asked the AI, what it got wrong, and how we checked.

Our research reaches 17,000+ followers across our channels. Upstream is the weekly version.

Weekly. Unsubscribe anytime. Privacy