Auto Merge
Auto merge is a way to automatically merge duplicates in a CRM: new duplicate records are found and merged on a schedule, or as soon as they appear, using the criteria of a duplicate search that has already been checked. Each merge follows the same primary record and merge rules a person would apply by hand.
It's the last step in automating deduplication. Automatic detection finds new duplicates and alerts a person, who then merges them. Bulk merging clears many sets in one go when someone starts it. Auto merge keeps merging the new duplicates without anyone starting each run. Because a merge is often hard to reverse, it depends on guardrails: only proven rules run, risky matches are held back, and uncertain ones go to a person for review.
See how Dedupely merges new duplicates on a schedule in your CRM
What Is Auto Merge?
Auto merge, also called automatic merging or scheduled merging, lets a trusted duplicate search merge the new duplicates it finds without a person starting each merge. It works within one object at a time: contacts, companies and deals in HubSpot; accounts, contacts and leads in Salesforce; people and organizations in Pipedrive; leads in Close; and contacts in Nimble and Google Contacts.
A simple example: a contact, J. Smith at jsmith@acme.com, is already in the CRM, and a web form creates a second J. Smith with the same email. On the next run, a search that matches on email and last name pairs the two, the merge rules choose which values stay, and one record remains. Another set, where a dozen records share "test@test.com," is held back instead, because a set like that points to bad data rather than one person.
Tools run auto merge in a few ways. Scheduled runs merge at set times, such as every night or every hour, which keeps merges predictable and easy to check. Continuous runs scan and merge as soon as new duplicates appear, which keeps the time a duplicate stays in the CRM short. Merge-on-entry rules act when a record is created or imported, which stops a duplicate before anyone works on it. All three rely on the same thing: matching rules that have already shown they pair the right records.
How Auto Merge Works
An automatic merge repeats a set of decisions that were made once, ahead of time:
- Start from a trusted search. The fields, match methods and filters come from a search that has been run and checked by hand, so each automatic run repeats a known result on new records.
- Set what each merge keeps. The primary record and the field rules are set before the first run, because no one previews each merged record.
- Choose when it runs. Options often include fixed intervals, set days, a time window outside working hours, or a continuous loop.
- Hold back risky matches. Safety rules exclude sets that are very large, built on placeholder or example data, or likely to hit a CRM limit. HubSpot, for example, caps the associations one record can hold and limits how many times a record can be merged.
- Route uncertain matches to review. Matches with low confidence go to a review queue, where a person decides whether to merge, reject or leave them.
- Merge and record the result. Clear matches merge, activity and associations move to the surviving record, and each merge is logged so it can be traced later.
A way to stop a run matters too. Stopping halts the merges still to come, but the ones already finished stay merged, so it limits a mistake rather than reversing it. That's why the rules and guardrails are set before the first run, not after.
Why Auto Merge Matters for CRM Teams
New duplicates keep arriving after a cleanup. Web forms, list imports, integrations and reps adding records by hand all create them, so a CRM that was clean last quarter drifts back toward split records. Each new duplicate counts the same customer twice in reports, splits activity between two timelines and can route one lead to two reps. In CRMs that bill by contact or record count, it's also a record paid for twice.
Finding those duplicates is only half the work. If a person has to open and merge each new set, the cleanup becomes a recurring task that competes with everything else, and sets that wait longer stay in reports and sequences longer. Automating the clear cases keeps the CRM close to its cleaned state between reviews, and leaves people to spend their time on the sets that need a decision.
The same automation raises the stakes of a bad rule. A manual merge mistake affects one set; an automatic rule repeats its mistake on every run until someone notices. A loose field, a missing merge rule or a placeholder value can merge records that belong to different people, and many merges are hard to reverse. That's why it pays to automatically merge duplicates only with tested rules, safety limits and a review route for the uncertain matches.
Getting Started With Auto Merge
Clean the existing backlog first, by hand or in bulk, and automate only a search that has proven itself. A good candidate is a search whose sets you would merge without changes, run after run. A search that still turns up false matches needs stricter fields before it runs on its own.
Set the merge rules and safety limits before you turn anything on. Decide which record survives and which value wins in each field, keep large sets and placeholder data out of the automatic runs, and decide where uncertain matches go. Start with a narrow search on strong fields, such as email, and a schedule you can easily check, such as once a day.
Watch the first few runs closely. Compare a sample of merged records in the CRM with the originals in the merge history, and look at what was held back or sent for review. Once the results hold up, run the search more often or automate the next one, and keep a way to stop a run if something looks wrong.
In Dedupely, Auto Merge runs on one Duplicate Search at a time, on the schedule you choose, and each merge follows your Primary Record and Merge Rules. It skips matches excluded by Safety Settings and sets already in Manual Review, and sends low-confidence matches to Manual Review instead of merging them.
Key Takeaways: Auto Merge
Auto merge automatically merges new duplicates on a schedule, or as soon as they're found, using the criteria of a search that has already been checked. It keeps a CRM close to its cleaned state as forms, imports and integrations add new duplicates. Because merges are hard to reverse and an automatic rule repeats its mistakes, it needs proven rules, merge rules set in advance, safety limits and a review route for uncertain matches. In Dedupely, Auto Match and Auto Merge are set per Duplicate Search: Auto Match notifies you of new matches, and Auto Merge merges them on your schedule. Auto Merge skips matches excluded by Safety Settings, sends low-confidence matches to Manual Review, and Merge History records each merge.
Frequently Asked Questions About Auto Merge
How do you automatically merge duplicates in a CRM such as HubSpot or Salesforce?
Start from a duplicate search you trust, set what each merge keeps, add safety limits, and then put the search on a schedule. The same steps apply in HubSpot, Salesforce, Pipedrive or Close, and to contacts, companies, leads or deals. Run the search by hand first and merge its results until the sets it finds need no changes. Then set the primary record and field rules, hold back large sets and placeholder data, and send uncertain matches to a person instead of merging them. Clear matches can then merge without manual review, while the doubtful ones still get it. In Dedupely, you turn on Auto Merge for a Duplicate Search and choose its schedule. Before it starts, a confirmation shows the search, matching, filters and schedule, and you acknowledge that it will merge all current and future duplicates matching the search and that merges can't be undone. Auto Merge skips Excluded matches and sets in Manual Review, and sends low-confidence matches to Manual Review instead of merging them.
Is it safe to merge duplicates automatically?
It's safe when the rules have been tested, the merge settings are in place and risky matches are kept out. The risk isn't the automation itself but repeating a weak rule on every run: a loose name match or a shared placeholder email can merge records that belong to different people, and a merge is often hard to reverse. Automate only searches built on strong fields, check the first runs closely, and keep a review route for anything uncertain. In Dedupely, merges can't be undone. Safety Settings are on by default and exclude risky matches from Auto Merge, such as matches with more than 20 records (the default) or built on poor or example data, and they show in the Excluded tab on the next scan. Auto Merge also skips matches set to "Always ask" in merge choices, since it can't ask. "Cancel all" stops a running Auto Merge, and merges already finished stay as they are.
Which duplicate searches should you automate?
Automate the searches whose results you already merge without changes, and keep the rest manual. Those are often strict searches on strong identifiers, such as an exact email address or a customer ID, combined with a second field like last name. Searches that use looser methods on names, or that compare fields many records share, find more duplicates but also more false matches, so they're better kept for review. A useful test: if you've run a search by hand several times and haven't had to reject or adjust a set, it's a candidate. If you're still rejecting pairs, tighten the fields first. In Dedupely, Auto Match and Auto Merge settings belong to one Duplicate Search, so you can automate a trusted search and leave the others for review. Auto Match notifies you when new matches reach a count you set, and Auto Merge merges them on a schedule. Dedupely Intelligence can't be automated.
How often should an automatic merge run?
Run it as often as new duplicates arrive and as often as you can check the results. A CRM that takes in many form fills or imports each day may suit hourly runs or a continuous loop, while one that changes slowly may need only a daily or weekly run. Start with a schedule you can easily check, such as once a day outside working hours, and run it more often once the results hold up. A run that overlaps with a large import or a sync from another system can be worth moving to a quieter time. In Dedupely, Auto Merge can run as soon as possible, in a continuous loop of scan, merge and scan again, or every set number of minutes, hours, days, weeks or months. It can also run on chosen weekdays or days of the month, and between chosen hours. Dedupely refreshes from the CRM every 10 to 30 minutes and detects new records within 1 to 5 minutes.
What happens to uncertain matches when duplicates merge automatically?
Uncertain matches should be set aside for a person to review, not merged. A good automatic setup sorts matches into three groups: clear ones that merge, risky ones that are held back, and uncertain ones that wait in a review queue. That way automation handles the obvious duplicates while a person decides the borderline sets, such as two records with the same name and a similar but not identical email. Check the review queue on a regular schedule, so held sets don't wait too long. In Dedupely, Auto Merge sends low-confidence matches to Manual Review instead of merging them, and skips sets already there. From Manual Review, you can merge a set or remove it. Matches that trip a Safety Settings rule go to the Excluded tab, where you can merge them by hand, reject them or send them to Manual Review. Rejected Matches keeps a rejected set out of current and future Duplicate Searches.
