What is data hygiene?

What does data hygiene include?

Data hygiene includes four routines, each aimed at one way records go wrong: deduplication, standardization, completion, and refresh with verification.

• Deduplication: one company or one person, one record. Merge on stable keys such as the domain, the company page or the email address, never on the name alone.
• Standardization: one format per field. “VP Sales”, “Vice President, Sales” and “VP of Sales” become one title; “USA”, “U.S.” and “United States” become one country.
• Completion: fill the empty fields, such as industry, size or country, from an outside source. That step has its own name: data enrichment.
• Refresh and verification: check that what you hold is still true. Companies rename themselves, people change jobs, mailboxes close. An address should go through email verification before a campaign, not after the bounces.

Why does CRM data go bad?

CRM data goes bad because the world keeps moving after the import. The example above shows it on a real company: a row that still says Facebook leads to the right company, because the company page, now named Meta, still lists facebook.com. Only the name in the CRM is out of date.

The same drift hits every field that describes something alive, such as titles, employers, headcounts, offices and email addresses. That slow drift is called data decay.

Bad data also comes in through the door. Event lists get imported twice, forms collect typos, and a rep creates an account that already exists under another spelling. Hygiene has to handle both: the data that ages, and the data that arrives broken.

Which is which?

Six problems found in the same CRM. Which kind is each one?

1. Two accounts, Acme Inc. and ACME Corporation, both on acme.com

2. An account still named Northwind Traders after the company renamed itself Northwind Health

3. An account with a domain and nothing else: no industry, size or country

4. A contact whose title still says Sales Manager, a year after a promotion

5. A lead with a work email and no company attached

6. The same contact imported twice, once from a webinar list and once by hand

Show the answers

1. Two accounts, Acme Inc. and ACME Corporation, both on acme.comDuplicateOne company held twice: merge on the shared domain and keep the newest values.
2. An account still named Northwind Traders after the company renamed itself Northwind HealthOutdatedThe value was right once. The company changed; the row did not.
3. An account with a domain and nothing else: no industry, size or countryIncompleteNothing is wrong yet, but fields are missing: enrichment fills them from the domain.
4. A contact whose title still says Sales Manager, a year after a promotionOutdatedTitles and employers drift as people move; only a refresh catches it.
5. A lead with a work email and no company attachedIncompleteThe email domain holds the answer: resolve it and link the lead to its account.
6. The same contact imported twice, once from a webinar list and once by handDuplicateTwo rows for one person split the history; match on the email address and merge.

Data hygiene vs data cleansing: what is the difference?

Data hygiene and data cleansing overlap, and teams use the two terms loosely. One common way to separate them: cleansing is a project that clears a backlog, hygiene is the routine that keeps the backlog from coming back.

	Data cleansing	Data hygiene
What it is	A project	A routine
When it runs	Often once, before a migration or a campaign; some teams schedule it	On a schedule, and on events such as an import
What it fixes	The backlog of duplicates, broken formats and dead records	New errors and new decay, as they appear

Cleansing is the catch-up; hygiene is the habit that keeps you from needing another one. Many teams run one cleansing project when the CRM becomes hard to use, then put hygiene in place so that the next one never comes.

How do you clean up CRM data?

To clean up CRM data, start with the keys, then work outward: decide how records match, stop bad formats at the door, merge on the keys, fill the fields that matter, and check again on a schedule. Entity resolution depends on the keys you choose.

1. KeysChoose how records matchDomain and company page for accounts; email or profile URL for contacts
2. At the doorStandardizePicklists in forms, rules on imports, one format per field
3. MergeMerge on the keysKeep the newest value and a note of the old one, such as "formerly Facebook"
4. FillFill the gapsIndustry, size, country and owner: the fields routing and scoring need

Check again on a schedule: monthly for accounts in active deals, quarterly for the rest, and verify emails before every campaign

How do you automate data hygiene with an API?

To automate data hygiene with an API, run the checks from the CRM itself, on a schedule or when a record changes.

Company Enrichment returns the current version of an account from its company page, so a weekly job can compare names, sizes and headquarters with what the CRM holds and flag the differences. When a row only has a domain, Domain to LinkedIn resolves it first, as in the example above. The CRM hygiene use case shows the full routine, including finding the current decision makers at each account and checking old emails again.