Data hygiene
Definition
The state of the records underneath your reporting: duplicate contacts, dead email addresses, empty required fields and free text where a picklist belongs. A report can never be more credible than the database it is built on.
How you recognize it
Two factories produce most of the mess. Forms that write a new record on every submission, so the same person exists once per whitepaper, each copy holding a different slice of the history and none holding all of it. And imports run without deduplication: a trade-fair list or an old spreadsheet lands on top of the live base, bringing its own date formats and its own spellings of every company name. Decay does the rest, because people change jobs and their addresses die behind them. What changes the pace is volume: every new form and every import either matches records or multiplies them.
What to watch for
Start with the source field: filter your contacts on source empty or unknown and read the share, because every contact in that filter is missing from any answer about which channel produced revenue; this one empty field is what makes attribution reporting unusable. Search for your own company and count the records that come back. Open the country or industry field and count the spellings of the same value. And read what people typed into the fields your forms force: a required field the visitor finds irrelevant collects dashes and junk, which then feed the segmentation your campaigns run on.
The question you ask
“How many of our contacts have no source on them?”