Duplicate contacts, and how to stop making them

Every CRM with more than one user accumulates duplicates, and the cause is almost always the same. Somebody searched for a customer, did not find them, and created a second record rather than hunting further. The search failed because of a spelling, an abbreviation or a business name entered three different ways. From the person's point of view the customer simply was not there.

That framing matters, because most businesses treat duplicates as carelessness. They are better understood as feedback about the search. A business that fixes the search stops producing them, and a business that only cleans up keeps producing them at the same rate.

What a duplicate actually costs

The history splits first. Two jobs sit on one record and three on the other, so neither answers what this customer has bought from you. A conversation about their last visit becomes a conversation about half their last visits. Nobody notices, because each record looks complete on its own.

The reporting goes next. A customer counted twice inflates the customer count and halves the average value. A mail merge sends two copies to the same office. An invoice goes to whichever address sat on the record somebody happened to open. None of those failures announces itself, which is why the count matters more than any individual example.

Fix the search before cleaning the list

A search that only reads the name will keep missing the customer whose business name somebody shortened. Acme, Acme Pty Ltd and Acme Plumbing are three strings and one customer. The phone number, by contrast, rarely gets typed a dozen ways. Two records sharing a mobile number are almost always the same person.

Diract CRM searches the phone number, the email and the address alongside the name. That finds the record somebody would otherwise recreate. An archived contact still appears, clearly marked, because a result of nothing found is the single most reliable cause of the next duplicate.

Merge, and decide which record survives

Deleting the emptier duplicate destroys whatever sat on it, and that is usually the job somebody cannot find later. Merging in Diract CRM keeps both histories against one record. Our guide to cleaning up duplicates covers doing that without losing the links each record carried.

The choice of survivor is worth stating as a rule. The older record usually holds more history, and the newer one usually holds the current phone number. Keeping the older record and taking the newer details onto it loses least. A rule written down once stops each merge becoming a small judgement call.

Imports are where duplicates arrive in bulk

A spreadsheet of four hundred contacts imported into a list already holding three hundred produces a mess nobody unpicks by hand. The duplicates arrive all at once, and they arrive with partial histories attached. Matching on email or phone before the import is the only version of this that stays manageable. Diract CRM matches on both during the import, which is where the problem is cheapest to solve.

A sample of twenty records first will reveal every mapping mistake the full file would make. That check takes a minute. Skipping it converts a minute into four hundred corrections, and usually into a decision to abandon the import halfway.

A naming rule and an owner

Half the duplicates in any list come from three people typing one business three ways. One agreed rule removes that entirely. The legal name on the invoice works well, because somebody can always check it. A single sentence records the rule, and every new starter reads it in their first week.

Data quality with no owner degrades, because each individual mess is small and nobody's job. A named person doing twenty minutes a month keeps a list of a few thousand contacts clean indefinitely. The same person should watch the count, because a falling number of repeated phone numbers is the proof the search fix worked.

What to change first

• A search reads phone, email and address as well as the name
• A monthly check looks for repeated phone numbers and emails
• Duplicates merge rather than getting deleted
• The older record survives and takes the newer details
• An import matches on email or phone, after a sample of twenty
• One agreed naming rule covers business names
• One named person owns the contact list