Contact management goes wrong in a predictable way. Everybody agrees there should be one list, nobody agrees who owns it, and within a year there are three — the shared one, the one in somebody's phone, and the spreadsheet that is actually current. The fix is rarely a better tool; it is deciding what the list is for and who is allowed to be wrong in it.
What contact management should record
- Name, organisation and role, with the role dated — people move, and a stale title is how an email reaches the wrong person.
- One primary channel that somebody actually answers, rather than four addresses of unknown freshness.
- The relationship: who in your team owns it, and why this person is on the list at all.
- Consent and source, where you contact people for marketing — this is a legal requirement in most of the world, not an optional field.
- The last meaningful interaction, which is the only reliable signal of whether a contact is live.
What not to record
Personal detail that serves no business purpose, opinions about people, and anything you would not want the contact to read. Records are disclosable, get exported, and outlive the person who typed them. The working test is whether you could show the entry to the contact without embarrassment — it costs nothing to apply and removes the entries that turn a database into a liability. Separately, resist custom fields that only one person uses: every unused field makes the record slower to fill in and less likely to be filled in at all.
Decide your decay policy before the list is big. A contact nobody has touched in two years is noise, and marking it dormant rather than deleting it keeps the history while removing it from everything that counts.
Stopping the duplicates
- Name one owner for the list. Shared ownership is the root cause of every other problem here.
- Agree a single identity rule — email address is usually best — and enforce it at entry rather than cleaning up later.
- Import once, deliberately, rather than letting everyone sync their phone into it.
- Run a duplicate check on a schedule, and merge rather than delete so history survives.
- Record consent at the point of capture; retrofitting it is impossible and auditors ask.
When a list becomes a CRM
The boundary is not size, it is whether you need to track a process. A list answers who somebody is and how to reach them. The moment you need stages, next actions and a forecast, you are describing a pipeline, and keeping that in a contact list produces the familiar mess of status words typed into a notes field. Moving earlier than you need to is the opposite mistake — a sales pipeline with four deals in it is overhead.
Ettex Contacts covers the list properly: people with organisation, owner, consent and source, the last interaction, and a dormant state rather than deletion. Where the work is a process rather than a record, the same people appear in a customer database or a pipeline without being retyped.
Frequently asked
Should contacts live in the CRM or separately?
One source of truth either way. The question to answer first is who owns it, because duplicates come from unclear ownership far more often than from the wrong tool.
How often should a contact list be cleaned?
Quarterly for duplicates, annually for dormancy. A list nobody prunes stops being trusted, and an untrusted list is immediately copied into a private spreadsheet.
Do I need consent records for business contacts?
In most jurisdictions yes for marketing, with narrow exemptions that vary. Record the source and date at capture — reconstructing consent later is not possible.