Manual Deduplication
Manual deduplication is the process of finding and merging duplicate CRM records by hand: a person searches or sorts the records to spot duplicates, compares each set and merges it, one set at a time. It's also called manual merging or merging duplicates one by one, and it's how many teams first merge duplicates manually before they use rules or automation.
Manual deduplication puts each decision in a person's hands, which suits a small number of duplicates and records where a wrong merge would be costly, such as key accounts or customers tied to billing. Its cost is time. Each set needs someone to find it, check it and merge it, so the work grows with the database, and duplicates that differ by a typo or a nickname are easy to miss when records are scanned by eye. Teams that weigh manual vs automated deduplication often end up using both: rules for the clear sets, a person for the rest.
See how Dedupely merges duplicates by hand and in bulk in your CRM
What Is Manual Deduplication?
Manual deduplication, also called manual merging, covers both stages of deduplication done by a person: spotting which records are the same, and merging them. It applies to every object a CRM stores: contacts and companies in HubSpot, accounts and leads in Salesforce, people and organizations in Pipedrive, leads in Close, and contacts in Nimble and Google Contacts.
A common way to start is to sort a contact list by email or last name and look for neighbors that match. Sorted by email, "Maria Garcia, maria@acme.com" and "M. Garcia, maria@acme.com" sit side by side and are easy to spot. A third record, "Maria García, mgarcia@gmail.com, Acme," lands further down the list and stays unmerged, because nothing in the sorted column links it to the other two.
Teams do the work in a few ways. Working inside the CRM, a person searches for a name, or opens the CRM's list of suspected duplicates where there is one, and merges from the record. Its strength is that nothing leaves the CRM. Working from an export, a person sorts and flags duplicates in a spreadsheet, then merges the flagged records in the CRM. Its strength is seeing many records at once and sorting by several columns. A third way mixes hand work with rules: a tool finds the duplicate sets, and a person checks and merges them one at a time. Its strength is that the finding no longer depends on someone's eye.
How Manual Deduplication Works
Manual deduplication follows the same steps as any deduplication, with a person doing each one:
- Pick the object and the fields. Start with one object, such as contacts or companies, and the fields most likely to reveal duplicates: email, phone, company name or website domain.
- Find the candidates. Sort, filter or search the records on those fields, or open the CRM's duplicate list where it has one. Each pass on one field finds the records that share it and misses the ones that differ by a typo, a nickname or a different format.
- Check each set. Compare the records side by side to confirm they describe the same person, company or deal. Owners, stages and activity history help, not only the field that matched.
- Choose what survives. Decide which record stays and which value each field keeps, and watch for blanks and placeholders such as "N/A" so a real value isn't lost.
- Merge. Built-in merge screens often take a few records at a time. Salesforce's native merge stops at three records, and Pipedrive merges duplicates one by one, so a larger set takes several merges.
- Check and note the result. Confirm in the CRM that activity and associations moved to the surviving record, and note what changed, since a merge is hard or impossible to reverse in many systems.
Each set repeats the last four steps, so the time grows with the number of sets, not the number of searches. New duplicates from forms, imports and syncs start the cycle again.
Why Manual Deduplication Matters for CRM Teams
Manual deduplication keeps a person in charge of each merge, and for some records that's worth the time. A key account, a customer with open deals or a record tied to billing or another system can affect invoices, routing and reports if it's merged wrongly. A person who checks the owners, stages and activity can catch problems that a field comparison misses.
At scale, the same approach has costs that show up in the data. Duplicates stay in the CRM while they wait to be found and merged, so contact counts stay inflated, reports count the same customer twice and activity history stays split across records. In CRMs that bill by contact or record count, each one also adds to the bill. Duplicates that only a looser comparison would find, such as "Sean" and "Shawn" or one phone number saved in two formats, often stay unmerged because nobody spots them.
Consistency is the other cost. When several people merge by hand, they make different choices about which record survives and which value wins, so merged records follow different logic. And because new duplicates keep arriving, a manual cleanup often has to start over. That's why many teams keep manual deduplication for the sets that need a person and use rules for the rest.
Getting Started With Manual Deduplication
Start by sizing the problem before choosing how to work. Run a few searches on strong fields, such as email or website domain, and see how many duplicate sets they find. A handful of sets can be merged by hand, while hundreds call for rules, with hand merges kept for the exceptions.
Then sort the sets by risk. Clear sets matched on several strong fields can follow one set of rules for the surviving record and each field. Sets tied to key accounts, open deals or billing, sets with conflicting owners or stages, and sets with unusual data are the ones worth merging by hand.
For the hand merges, work through one search at a time and decide each set fully: compare the records, choose the values, merge and check the result in the CRM. Keep a record of what each merge changed. Merging a few typical sets by hand before a large merge also shows whether the rules keep the values you'd choose.
In Dedupely, Custom merge carries out the hand merges: you open a duplicate set, pick the winning value field by field, preview the result and merge. The clear sets go through Bulk Merge, which merges several sets, a whole page or all the matches at once with the Primary Record and Merge Rules you've set.
Key Takeaways: Manual Deduplication
Manual deduplication is finding and merging duplicate records by hand, one set at a time. It gives a person control over each merge, which suits a small number of sets and records where a wrong merge would be costly. The trade-offs are time that grows with the database, near-duplicates that are easy to miss by eye and different choices from different people, so many teams keep it for the exceptions and use rules for the rest. In Dedupely, Duplicate Searches find the sets, Custom merge lets you pick the winning value field by field, and Manual Review holds uncertain sets for a person. Bulk Merge merges the clear sets with your Merge Rules, Auto Match notifies you of new matches, Auto Merge merges new duplicates on a schedule, and Merge History records each merge's original records, since merges can't be undone.
Frequently Asked Questions About Manual Deduplication
How long does it take to merge duplicates manually in a CRM?
It depends on how many duplicate sets there are, because each set is its own round of finding, checking and merging. A few sets fit in a short session. A large backlog takes much longer, since every set repeats the same steps and built-in merge screens often take only a few records at a time. The time also grows as the database grows and new duplicates arrive from forms, imports and syncs. That holds in HubSpot, Salesforce, Pipedrive or another CRM. Two other costs come with the time: duplicates stay in the CRM while they wait, and in tools billed by contact count, each one is a record the company pays for twice. In Dedupely, a Duplicate Search finds the sets, and Bulk Merge merges several sets, a whole page or all the matches in one action, with the Primary Record and Merge Rules you've set applied to each one. In Salesforce, it can merge thousands of records that way.
When is manual deduplication the right choice?
Manual deduplication is the right choice when there are only a few duplicate sets, or when a wrong merge would be costly. That covers key accounts, customers with open deals or contracts, records tied to billing or another system, sets where the records disagree on owners or lifecycle stages, and sets with unusual data that rules don't handle well. Merging a few typical sets by hand is also a good way to check merge rules before running them on many sets. For a large backlog of clear duplicates matched on strong fields, such as an exact email, rules are faster and more consistent. In Dedupely, Custom merge is the manual way to merge: you open a set with View Match Details, pick the winning value field by field, preview the result and merge. Manual Review holds sets that need a person's decision until they're merged or removed, and matches held back by Safety Settings in the Excluded tab can still be merged by hand.
What's the difference between manual and automated deduplication?
Manual deduplication has a person find and merge each duplicate set, while automated deduplication uses rules to find the sets and merge them in bulk or on a schedule. Manual work gives full control over each merge and suits a small number of sets or high-stakes records. Automated deduplication suits volume: it compares many records on several fields at once, can catch near-duplicates such as typos and alternate spellings, and applies the same rules to each merge, so the results are consistent. Its risk is that a bad rule repeats in each set, so rules are checked on a sample first and uncertain sets are held back for a person. Many teams use both. In Dedupely, Duplicate Searches find duplicates with a Match Option on each field, Merge Rules decide what each merge keeps, and Bulk Merge merges many sets at once. Auto Match notifies you when new matches reach a count you set, and Auto Merge merges new duplicates on a schedule.
How can you make merging duplicates manually faster?
Let a search do the finding, and keep the hand work for the sets that need it. Scanning records by eye is often the slowest part, and it misses duplicates that differ by a typo or a nickname. Search on strong fields, such as email, phone or website domain, and combine a few of them so each result is likely a real duplicate. Then set rules for which record survives and which value each field keeps, merge the clear sets together, and merge by hand only the sets that look unusual or matter most. Work one search at a time so similar sets are handled the same way. In Dedupely, preset Duplicate Searches for common field combinations are ready the first time you log in. Merge Rules can suggest rules per field, such as "Placeholders lose" or "Primary or best phone," and a preview shows the merge result while you build a rule. Bulk Merge then merges the clear sets in one action.
How is manual deduplication different from manual review?
Manual deduplication is doing the whole job by hand, while manual review is one step in a deduplication process where uncertain sets wait for a person's decision. In manual deduplication, a person finds the duplicates, checks them and merges each set. In manual review, a search or rule has already found the sets, and the clear ones can be merged in bulk or on a schedule; only the uncertain ones wait in a queue for a person to merge or reject. Manual review keeps a person's check where it's needed without making the whole cleanup manual. In Dedupely, Manual Review is that queue: you send a duplicate set there from any Duplicate Search, and it stays until it's merged or removed. Auto Merge and Bulk Merge leave out sets in Manual Review. Two routes send sets there automatically: Auto Merge's low-confidence matches, and a per-field "Send incorrect formats to manual review" checkbox for invalid and filler values.
