What is data enrichment?

How does data enrichment work?

Data enrichment takes an identifier your record already holds, matches it to the right company or person in an outside source, and appends the fields you asked for. The match is the hard part: a domain can be claimed by several company pages, and a name by thousands of people.

1. Start fromAn identifier you holdA company domain, a company or profile URL, or an email address
2. MatchThe right entityThe one company or person the identifier points to, in the source
3. AppendThe fields you asked forEach with the value the source holds on the day of the call

The Figma example above runs those steps on a CRM row that held only a domain. One call matches the domain to the company, a second returns the company record, a third counts its open job postings.

What can data enrichment add to a B2B record?

Data enrichment can add four kinds of data to a B2B record, each from its own starting point:

Kind of data	Examples	Typical starting point
Company data	Industry, size, headquarters, founding year, email provider	A domain or a company page
People data	Current role, seniority, employer, work history	A profile URL or a work email
Contact data	Verified work email, direct phone number	A profile URL
Signals	Open job postings, tools named in them	A company page

Company data on its own has a name in the glossary: firmographic data. The tools a company uses are its technographic data.

Why do records need to be enriched again?

Records need to be enriched again because the facts they hold change while they sit in the CRM. People change jobs and titles, companies hire, move, merge and rebrand, and email addresses stop working when someone leaves. An enriched record is true on the day of the call; after that, it drifts. That is why enrichment runs twice: once when a record enters the CRM, and again on a schedule.

Data enrichment, data cleansing and waterfall enrichment: what is the difference?

Data enrichment adds what a record is missing, cleansing fixes what it already holds, and waterfall enrichment asks several sources in turn for the same missing field.

	What it does	When it runs
Data cleansing	Fixes what is already there: duplicates, typos, formats	Before matching; often a one-off project, sometimes on a schedule
Data enrichment	Adds what is missing, from an outside source	When a record enters the CRM, and on a schedule
Waterfall enrichment	Asks several sources in turn for the same missing field	When one source alone does not find the field

Which is which?

Five jobs on the same CRM. Which one is each?

1. Adding industry, size and headquarters to a list of domains

2. Merging two contacts that turn out to be the same person

3. Asking a second source for the work emails the first one did not find

4. Rewriting "VP, Sales" and "Vice President Sales" as one title

5. Getting a contact's current role from their profile link

Show the answers

1. Adding industry, size and headquarters to a list of domainsEnrichmentNew facts come from an outside source and are added to rows you already hold.
2. Merging two contacts that turn out to be the same personCleansingNothing new comes in: the fix works on what the CRM already holds.
3. Asking a second source for the work emails the first one did not findWaterfallThe same missing field, asked of several sources in turn until one finds it.
4. Rewriting "VP, Sales" and "Vice President Sales" as one titleCleansingStandardizing a format fixes values that are already there.
5. Getting a contact's current role from their profile linkEnrichmentAn identifier you hold brings back a fact you did not have.

How do you enrich data with an API?

Enriching data with an API takes one call per identifier, from wherever the records live: a CRM, a spreadsheet, a workflow tool such as Clay or n8n, or an AI agent.

Company Enrichment turns a company page into a full record, Email Enrichment returns a verified work email from a profile URL, and the profile enrichment use case shows how the calls chain on a contact list. Each call returns structured JSON, so the fields land in your own columns without manual work.