Customer matching — which records are the same customer?

Customer matching answers one question: Which records across my tools are the same customer? Your billing system, your CRM and your other tools each keep their own record of a customer. Customer matching joins those records into one customer with one ID, shows you what it joined and why, and waits for a person on your team to confirm each match. Every other module counts customers on that ID. It belongs to the Shared foundation — see Modules.

​

What it does

As soon as two of your tools hold customer records, Beacon joins the records that clearly describe the same company — today, records with the same email domain or the same stored ID. Each customer then has one of two statuses:

• Matched — the customer is complete: found in every tool it belongs in, linked to its parent company if it has one, confirmed by a person, and checked recently.
• To review — something is still open, and the page says what: a match to confirm, a record to find in another tool, a parent company to link, or a re-check.

The aim is every customer Matched. A joined customer is used in your figures straight away, but it stays To review until someone confirms it. Beacon never merges two records on its own, and it never changes or merges anything in your tools — the matches are kept in Beacon.

​

What it produces

Its headline is one figure: It is shown on the module’s page in Beacon as not yet calculated. Its name links to its definition, which also says how the Data API and MCP will carry it — neither serves it yet. Behind it, the page shows today:

​

How a customer becomes Matched

The match score has four parts. A customer is Matched when all four are complete — its score is then 100. Each part is something a person can close: The weights are fixed. Your billing system and your CRM count double in the first part. When Beacon compares two records it looks at seven details: email domain, company name, billing address, contacts the two records share, tax or registration number, website and phone. Compare records shows each of them side by side. Today Beacon joins only exact matches; scoring looser, possible matches waits for the match score.

​

What you see on its page

In Beacon, open The Brain and choose Customer matching. The page has six tabs:

• Overview — customers matched and to review, the next reason to review, where each customer stands, the steps to 100%, and the tools it reads.
• Dashboard — the module’s main charts. The revenue charts read Not yet calculated today.
• Signals & alerts — what the module watches, and its alert rules.
• Scorecard — the four parts of the match score and their weights, the seven details and the rules. The scores read Not yet calculated.
• Customers — the review queue, Match by hand, Compare records and every customer across your tools. Confirm, keep two records apart, say a customer is not in a tool, or undo.
• Setup — the questions the module uses, and the answers in use.

Every confirmation, keep-apart and undo is recorded with the name of the person who made it. Confirming happens one customer at a time today; confirming every exact match in one go is not available yet. For a company with thousands of customers the page can take up to half a minute to open.

​

What it needs

Matching starts as soon as any two connected tools hold customer records. No single tool is required. See Connect your tools for how each connection is made.

​

Reads from

Nothing. Customer matching runs first, before any other module, and reads your tools directly. Later it will also take new email domains and segments from Segments, to spot new customers and show matching by segment.

​

Which modules read it

Today the other modules use the join itself. The ones that use the match score — Customer fit, Data integrity and Segment revenue paths — wait for it, because it is not calculated yet. The Modules page in The Brain shows which modules are running for you.

​

Alerts

No alert is sent yet. The first rule is built to post to Slack and waits for the match score; the others wait for the parts they read.

​

Setup questions

• Which tool is right about each customer’s details — name, address, tax number?
• How should a subsidiary’s revenue be counted? Recommended: all of it to the parent company, unless you say otherwise for a particular customer.
• Which customers count as your biggest? They are reviewed first, and their alerts come first.

Which tools you use and who looks after your data are asked once and used here too.

​

What is not calculated yet

• The match score for each customer, its four parts, and the tiers that tell other modules whether to use a customer still to review.
• Revenue matched and revenue by status.
• Parent companies and how a group’s revenue is counted.
• Possible matches — Beacon joins exact matches only.
• Confirming every exact match at once.
• A daily matching run, so new records are checked within a day.
• Product usage as a column, and the other tools in the table above.
• Alerts — none is sent yet.

​

Related