Blog
Team Collaboration

How to set up a review process for uncertain matches

7
min read
A stack of unreviewed matches under a cobweb at an empty desk

Table of contents

Talk to sales
Try for free
View as Markdown

TL;DR

  • Uncertain duplicate matches need one owner who runs the review and keeps the queue moving.
  • Splitting review by record owner works well when owners know their own accounts.
  • Every reviewed set ends one of three ways: merged, rejected or a changed search.
  • A regular review routine keeps uncertain matches from waiting for months.
  • Dedupely lets you send uncertain sets to review, reject false matches and get alerts.

Every duplicate search produces two kinds of results. Some sets you'd merge without looking, and some you aren't sure about.

The sure ones are easy to handle in bulk or on a schedule. The uncertain ones are where cleanups stall.

They sit in a list nobody owns, or they get merged in a hurry by whoever is closest. Neither is a good outcome, since a merge changes your live CRM.

In this post, you'll see how to set up a review process for uncertain matches. That covers who owns it, how to split the work, what each decision looks like and how often to review.

What makes a match uncertain

A match is uncertain when the fields that agree don't prove the records are one customer. Common examples are a shared phone, a similar company name with a different domain, or two records owned by different reps.

What counts as a duplicate is case by case. What one team considers a duplicate, another may not. That's why uncertain sets need a person who knows the business.

A review process answers three questions: who looks at these sets, how the work is split and what happens after each decision.

Give the review one owner

Some teams try to share review across everyone who uses the CRM. In practice, many end up with one person in charge, like a CRM admin, an ops lead or someone in RevOps.

The owner doesn't have to decide every set. Their job is to keep the queue moving, send each set to whoever can answer and make sure decisions get recorded.

Reps can still help with the decisions when they don't merge. They often know which of two accounts is real.

The owner just keeps the merge button in fewer hands. Our post on delegating a CRM cleanup covers what to agree on before someone else does the merging.

Split the review in a way that fits your team

By record owner

Some teams split by who owns the records in each set. When both records belong to the same rep, there's no ownership question, and the set can often be merged in bulk. When the records belong to different reps, the set goes to review, because someone has to decide who keeps the account.

For larger teams, one search per owner can give each rep a short list of their own accounts to check, merged one by one. When a rep finds that someone else owns one of their accounts, they raise it with the admin or the sales manager.

By type of doubt

Another way is to split by why a set is uncertain. Households and shared phones go to whoever handles customer data, related companies go to account owners and partner records go to the partner team.

By data owner, with a central team

In bigger companies, a central data team can run the searches while the business owners of the data choose which record survives. The central team keeps the process the same everywhere, and the business keeps the decisions. Some teams plan to hand more of the work to each business later.

Split Works well when Watch out for
One owner decides everything The team is small and one person knows the data A queue that grows faster than one person can clear
By record owner Reps know their accounts and ownership is clear Sets with two owners need a tiebreaker
By type of doubt Different teams own different kinds of records Sets that fit no category
Central team plus business owners Several business units share one CRM Slow answers from busy owners

Decide what happens after each review

Each reviewed set should end in one of three ways. Leaving a set "for later" is how review queues stall.

  • Merge it. The records are one customer. Check which values and owner the merged record keeps, then merge.
  • Reject it. You're certain the records aren't duplicates. A rejected match stays out of later searches, so nobody reviews it twice.
  • Change the search. The same false pattern keeps showing up, like households sharing a phone. Tighten the search instead of deciding each set by hand.

Write down decisions that set a pattern, like keeping branches with their own contract separate. A short list saves the next reviewer from reopening the question.

Decide whether you need a sign-off

Formal approval steps for merges are uncommon. In many teams, whoever runs the merge is also the one who decides it.

That works when the reviewer knows the data and the rules are written down.

A second approver makes sense for a few cases. Think of key accounts, records other systems depend on, or a team that's new to the data and wants IT involved before anything merges. Keep the sign-off for those sets rather than every merge, or the queue won't move.

Set a review routine

Uncertain matches keep arriving as new records come in. A routine keeps them from waiting for months.

  • During a cleanup: review the uncertain sets of each search before you move to the next one.
  • After the cleanup: check sensitive searches once or twice a month, or when their new matches reach a set number.
  • After events that add data: review again after a large import, a migration or a new integration.

Our post on how often to deduplicate your CRM covers cadence for the rest of your searches.

How Dedupely supports a review process

In Dedupely, you can send any set from any search to review, where it waits until someone merges or removes it. To split the work, you can build one search per owner with a filter on the owner field. You can also reject sets you're certain aren't duplicates, so they stay out of later searches.

Alerts can tell your team when a search's new matches reach the number you set, with at most one email per day.

Set up review for your uncertain matches

A process for the sets nobody is sure about

Uncertain matches need one owner, a split that sends each set to someone who knows the records, and a clear ending for every decision.

With a regular routine on top, the queue stays short. Every rejected match and tightened search also means fewer uncertain sets the next time around.

Frequently asked questions

Who should review uncertain duplicate matches?

One person should own the review, often a CRM admin, an ops lead or someone in RevOps. They send each set to whoever knows the records, and each decision should end in a merge, a rejection or a changed search. They also record decisions that set a pattern.

Should sales reps merge their own duplicates?

Reps are useful reviewers for their own accounts, but many teams keep merging in fewer hands. A common approach is to merge same-owner duplicates in bulk and send sets with different owners to review. A rep who finds an ownership problem raises it with the admin or a sales manager.

Do CRM merges need approval?

Usually not for every merge. Many teams let the person running the merge decide, as long as the rules are written down. A second approver makes sense for key accounts, records other systems depend on or cleanups where IT wants a say before anything merges.

How often should you review uncertain duplicates?

During a cleanup, review each search's uncertain sets before moving on. After that, check sensitive searches once or twice a month, or when new matches reach a set number. Review again after large imports, migrations or new integrations, since those add uncertain sets quickly.

What should you do with a match that isn't a duplicate?

Reject it once you're certain, so it stays out of later searches and nobody reviews it again. If the same kind of false match keeps appearing, change the search instead, for example by adding a field that tells the records apart.

Portrait of Gaby Salinas

Behind the article

Gaby Salinas

Marketing Assistant at Dedupely

Gaby Salinas came to CRM data from marketing and advertising, so she looks at duplicates from two sides: the match and merge logic behind them, and what they do to the people using the CRM, like the rep who calls a customer twice, the admin who gets blamed and the cleanup nobody wants to start.

As Marketing Assistant at Dedupely, she follows how RevOps teams, end users, consultants and technical teams actually work with their data: the records they stop trusting, the workarounds they build and the moment they finally decide to clean up. What she finds shapes most of what Dedupely publishes.

Connect with Gaby on LinkedIn

Contact us

We’d be happy to help you get this set up.

Write us a message

We probably know the answer to your question already

Email copied to your clipboard!

Book a Zoom

Whether you’re getting started or getting intense.
Get in touch!

Get started for free

20 free trial merges and as much free support as you need to get your duplicates under control.

"It couldn't have been much easier to connect to HubSpot and run my first dedupe."

Emily K

Mercy Housing

HubSpot MarketplaceHubSpot user

Running CRM cleanup for clients? See the Partner Program

Skip to content