Custom Merge
Custom merge is a way of merging one duplicate set by hand, field by field: a person compares the records side by side and chooses the value each field keeps before the records are combined. It's also called a manual merge or a field-by-field merge, and it's the most direct way to choose which values to keep when merging.
Merge rules apply the same instructions to many duplicate sets at once. A custom merge puts each choice in a person's hands, one set at a time. That suits records where a wrong value has real consequences, such as a key account or a contact tied to billing, sets with unusual data that rules don't cover well, and checks on whether a set of merge rules picks the values a team expects. Its cost is time. Each set needs a person's attention, so a custom merge fits a handful of sets, not a large backlog.
See how Dedupely merges duplicates field by field in your CRM
What Is a Custom Merge?
A custom merge is a merge where a person, not a rule, chooses which values to keep when merging. It works on any object that holds duplicates: contacts in HubSpot, Nimble and Google Contacts, accounts and leads in Salesforce, people and organizations in Pipedrive, and leads in Close.
Take two records for Dana Lee at Acme. The older one has the work email dana@acme.com, the lifecycle stage Customer and the job title VP Sales. The newer one came in from a webinar form, with a personal email, a new mobile number, the job title Chief Revenue Officer and the lifecycle stage Lead. A rule that keeps the most recent value would move a paying customer back to Lead. In a custom merge, the reviewer keeps the work email and the Customer stage from the older record, and the new title and mobile number from the newer one.
Tools handle the manual choice in a few ways. Some show every field side by side and ask for a choice in each, which gives full control and suits records where every value matters. Others start from the values the merge rules would pick and let the person change only the ones that look wrong, which is faster and shows how the rules behave. In both, the person sees the outcome before anything changes in the CRM.
How a Custom Merge Works
A custom merge follows the same decisions as any merge, with a person making each one:
- Open one duplicate set. The records in the set appear side by side, so differences in each field are easy to see.
- Pick the record that survives. Many tools ask which record stays as the primary record. It often keeps its ID, and it supplies any value the person doesn't change.
- Choose the value for each field. For each field where the records disagree, the person selects the value to keep. Blank fields and placeholders such as "N/A" or "Unknown" need a closer look, so a real value isn't replaced by an empty one. In fields that hold several values, such as additional email addresses in some CRMs, some tools can keep values from more than one record.
- Check what moves over. Activity such as calls, emails, notes and tasks, and associations such as linked companies, deals and contacts, move to the surviving record. Details vary by CRM, so it's worth knowing what yours carries over.
- Preview and merge. A preview shows the final record with the chosen values. Once it looks right, the merge runs and the other records are removed.
Because the person sees each value, a custom merge also shows where the merge rules would have chosen differently. Those differences point to the rules worth changing before the next bulk merge.
Why Custom Merge Matters for CRM Teams
Many duplicate sets follow patterns a rule can handle, but some don't. A contact who changed jobs, an account split between two regions, or a record where the newer value is the wrong one can all slip past a rule that works well everywhere else. When those sets merge on rules alone, the result can look clean while holding the wrong owner, stage or phone number, and the error is harder to spot than the duplicate was.
The records that need the most care are often the ones that matter most: key accounts, open deals, records tied to billing or contracts, and records another system points to. A wrong value there affects reports, routing and the next conversation with the customer. Many merges are hard to reverse, so a person's check before the merge costs less than fixing the record afterward.
Custom merges also make merge rules better. Merging a few typical sets by hand shows which values a team would choose, and where the rules pick something else. Fixing those rules before a large bulk merge keeps the same mistake from repeating across many records. The trade-off is time, which is why manual merging works best on a small number of sets.
Getting Started With Custom Merge
Start by deciding which sets deserve a person's attention rather than a rule. Records tied to billing, contracts, open deals or another system, sets with conflicting stages or owners, and sets with unusual data are good candidates. Clear duplicates matched on several strong fields can follow the merge rules instead.
For each set, choose which values to keep when merging, field by field. Check the fields that change often, such as job title, phone and lifecycle stage, and look for blanks and placeholders before accepting a value. Preview the final record before merging, and check in the CRM afterward that activity and associations moved over.
Use a few custom merges to check the merge rules before a bulk merge. Pick typical sets, choose the values by hand and compare them with what the rules would keep. Where they differ, adjust the rules, then let them handle the rest of the sets.
In Dedupely, Custom merge carries out this approach: you open a duplicate set with View Match Details, pick the winning value field by field, preview the result and merge. It's the most manual way to merge, which makes it useful for edge cases and for checking Merge Rules before they run on many sets.
Key Takeaways: Custom Merge
Custom merge is the manual, field-by-field way to merge one duplicate set: a person compares the records, chooses the value each field keeps, previews the result and merges. It suits sensitive records and edge cases, and it's a practical way to check merge rules before a bulk merge, at the cost of time on each set. In Dedupely, Custom merge opens a set with View Match Details so you can pick the winning value field by field, preview and merge. Merge Rules handle the repeatable cases automatically, Manual Review keeps uncertain sets for a person, and Merge History records each merge with the original records and their values. Merges can't be undone, so the preview is the last check.
Frequently Asked Questions About Custom Merge
When should you merge duplicates by hand instead of with merge rules?
Merge by hand when a wrong value would be costly or when a set doesn't fit the patterns your rules expect. That includes key accounts, open deals, records tied to billing or contracts, sets with conflicting owners or lifecycle stages, and sets where the newest value is the wrong one. Merge rules are the better choice for clear, repeatable cases, such as duplicates matched on several strong fields, because they apply the same logic to many sets without a person checking each one. A common split is rules for the bulk of the sets and manual merges for the few that need a closer look. In Dedupely, Custom merge is the manual option. Manual Review holds sets for a person to check, and Auto Merge sends low-confidence matches there instead of merging them. Matches held back by Safety Settings in the Excluded tab can still be merged by hand, rejected or sent to Manual Review.
What's the difference between a custom merge and merge rules?
A custom merge is a person choosing each value for one duplicate set, while merge rules choose the values automatically, the same way, for every set they apply to. Rules are repeatable and fast, so they suit bulk and scheduled merging. A custom merge is slower but gives full control over one set, so it suits edge cases and sensitive records. The two work together: rules handle most sets, and custom merges cover the exceptions and show where the rules need adjusting. In Dedupely, Custom merge means opening a set with View Match Details and picking the winning value field by field by hand. That's different from the Custom option inside Merge Rules, which sets the winning record or value automatically from Win Rules and Lose Rules, such as making a placeholder lose or a record with a certain lead stage win. Lose Rules run before Win Rules, and Win Rules run from top to bottom.
How do you test merge rules with a custom merge?
Pick a few typical duplicate sets, choose the values you'd keep by hand, and compare them with what the rules would choose. Include sets with blanks, placeholders, changed job titles and different lifecycle stages, since those are where rules most often pick the wrong value. Where your choice and the rule's choice differ, adjust the rule, and repeat with new sets until they agree. Then run the rules on a small batch and check the results in the CRM before merging the rest. In Dedupely, Custom merge is the most manual way to merge and is meant for checking Merge Rules and edge cases: you open a set with View Match Details and pick each value yourself. Merge Rules show the result before merging in the in-line preview, with winning values in bold, and in the Preview Merge panel, and a preview of the merge result appears while you build a rule. Suggested rules, such as "Placeholders lose," can be applied or dismissed per field.
How do you choose which values to keep when merging contacts in a CRM such as HubSpot?
Go field by field and keep the value that's most current and most trusted, rather than taking everything from one record. Recent values often win for fields that change, such as job title or phone, while values from a trusted source, such as a billing address from finance, should win over form entries. A filled value should beat a blank or a placeholder. The same approach applies in HubSpot, Salesforce, Pipedrive and other CRMs, though what each CRM carries over differs: HubSpot, for example, gives the merged record a new ID. For a few sets, choose the values by hand; for many, set merge rules. In Dedupely, Custom merge lets you pick the winning value in each field for one set, and Merge Rules set a Primary Record and a rule for each field across many sets. Without Merge Rules, Dedupely follows the CRM's own merge logic, and in any Dedupely merge, blank values never win. All activities and associations are preserved on merge.
Can you keep values from more than one record in a merged field?
Only in fields that can hold more than one value, such as additional email addresses in some CRMs. Most fields hold a single value, such as a job title, lifecycle stage or owner, so the merged record keeps one winner and the other values are dropped. If both values matter, look for a multi-value field the CRM already has, or move the second value into another field before merging. In a custom merge, the person picks the winner for each single-value field. In Dedupely, a Custom merge can keep values from more than one record in a multi-value field, and in Merge Rules, Append combines them automatically, but it's only available for the CRM's own multi-value fields, such as Additional email addresses in HubSpot. Single-value fields use Primary, Custom or, on some fields, a "Best" option such as Primary or best email. Merge History keeps every original record and its values, and its CSV download lets you restore a dropped value by hand, since merges can't be undone.
