# Duplicate Search – Definition, FAQs & How Dedupely Helps > Duplicate search explained: what it is, how saved matching criteria find duplicate records in your CRM, and how to layer and schedule searches over time. - URL: https://dedupe.ly/glossary/duplicate-search - Section: glossary - Site index: https://dedupe.ly/llms.txt - Any page: append `.md` to its URL for this plain-text form --- [Glossary](https://dedupe.ly/glossary) Duplicate Search ## Duplicate Search A duplicate search is a saved set of matching criteria that runs against a CRM and returns groups of records that describe the same person, company or deal. Each search names three things: the object to check, the fields to compare, and how strictly to compare each field. Because the criteria are saved, the same search can run again whenever new records come in, and it returns results on the same terms each time. A CRM often needs more than one: a strict search for the clearest duplicates, looser ones for near-matches, and separate searches for each object. The search itself only finds and groups records. Merging them is a separate step, done by a person or by an automated merge tied to the search. See how Dedupely's Duplicate Searches find duplicates in your CRM [Try for free](https://v2.app.dedupe.ly/signup) [Talk to sales](https://meetings.dedupe.ly/meetings/gaby-salinas/talk-to-sales) Overview A duplicate search is a saved set of matching criteria that finds duplicate records in a CRM. It names the object, the fields to compare and how strictly to compare each, then groups matching records into duplicate sets. Teams often layer several searches and rerun them on a schedule. On this page - What Is a Duplicate Search? - How a Duplicate Search Works - Why Duplicate Searches Matter for CRM Teams - Getting Started With Duplicate Searches - Key Takeaways: Duplicate Search - Frequently Asked Questions About Duplicate Search - Related Terms and Concepts See how Dedupely's Duplicate Searches find duplicates in your CRM [Try for free](https://v2.app.dedupe.ly/signup) [Talk to sales](https://meetings.dedupe.ly/meetings/gaby-salinas/talk-to-sales) ### What Is a Duplicate Search? A duplicate search is the working unit of record matching: one saved question about the data, asked the same way each time it runs. It goes by several names, including duplicate rule, matching rule, dedupe search and duplicate scan. Some CRMs offer a built-in version as part of their duplicate detection, and dedicated deduplication tools offer their own. Each duplicate search is built from the same three choices. The object decides which records are compared: contacts, companies or deals in HubSpot, accounts or leads in Salesforce, people or organizations in Pipedrive, leads or opportunities in Close, or contacts in Nimble and Google Contacts. The fields decide what counts as evidence, such as email, name, phone or website. The strictness decides how close two values have to be: identical, close despite a typo, alike in sound, or sharing a root domain. Change any of the three, and the results change with it. Take two contacts: "Dana Lee, dana@acme.com" and "Dana Lee, d.lee@acme.co.uk." A search that compares email exactly keeps them apart. A search that compares first and last name exactly, plus email by root domain, groups them into one duplicate set. Same records, different search, different result. Because each search only finds the pattern it describes, CRM teams often run several side by side, each written for one kind of duplicate. Resources: - [Match Options](https://dedupe.ly/knowledge-base/match-options) (KB) - [Filters](https://dedupe.ly/knowledge-base/filters) (KB) ### How a Duplicate Search Works A duplicate search follows the same cycle in any tool: define it, run it, review what it finds, and run it again as new records arrive. - **Define the criteria.** Pick the object, the fields and how strictly to compare each one. Combining a few fields, such as first name + last name + company, cuts down the false matches a single field like name produces on its own. Filters can narrow which records are checked, for example only records that have an email address. - **Run the search.** The tool compares records on the chosen fields and groups the ones that match into duplicate sets. Running a search reads the data; it doesn't change it. - **Review the results.** Someone checks the sets and merges the real duplicates. Many tools also let you mark sets that aren't duplicates, so they stay out of future results. - **Layer more searches.** Once the strictest search is clean, looser searches catch what it missed: a typo, a phone number in another format, a name that sounds alike. One search per pattern keeps each list easy to review, and changing one search leaves the others as they were. - **Run it again on a schedule.** New records keep arriving from forms, imports, sales tools and syncs, so searches are rerun on a schedule or as records come in. Depending on the tool, a scheduled search can notify someone when new matches appear, or merge them automatically. Resources: - [Finding Duplicates](https://dedupe.ly/knowledge-base/find-duplicates) (KB) ### Why Duplicate Searches Matter for CRM Teams Without saved searches, finding duplicates means building the same query again each time, and the result depends on who builds it and how. A saved search makes the definition of a duplicate clear and repeatable: the same records give the same matches, and anyone on the team can see why two records were grouped. How the searches are set up also decides how much risk each merge carries. One broad search mixes clear duplicates with doubtful ones, so the whole list has to be checked by hand or merged with mistakes in it. Separate searches, one per pattern, sort the work by confidence. Exact email matches can be merged in bulk or automated, while a search on similar company names in the same city goes to a person first. A false match then stays inside one search, where it's easy to find and fix. Searches that run on a schedule keep the effects of duplicates small. Contact counts stay accurate for reporting and for CRMs that bill by record count, two reps don't work the same lead, and activity history stays on one record instead of splitting across several. The longer a search goes without running, the more duplicates build up before anyone sees them. Resources: - [How to Approach CRM Deduplication](https://dedupe.ly/knowledge-base/how-to-approach-crm-deduplication) (KB) - [Why your CRM database is turning into a mess (and how to fix it)](https://dedupe.ly/blog/why-your-crm-database-is-turning-into-a-mess) ### Getting Started With Duplicate Searches Build your searches from strict to loose. Start with one search per object on the strongest fields, such as email for contacts or domain for companies, and check what it finds before merging anything. Then add a looser search for each pattern the first one misses, such as similar names with the same phone number, or similar company names on the same root domain. One search per pattern keeps each list easy to review, and the searches that prove reliable over a few runs are the ones worth running on a schedule. Check each search's results over time. A search that keeps returning sets that aren't duplicates needs a stricter field, and a search that returns nothing may be comparing a field that's often blank. Placeholder values, such as "N/A" or a shared info@ address, are worth filtering out before a search runs, because they match across unrelated records. In Dedupely, Duplicate Searches work this way: each search compares the fields you choose, with a Match Option for each one, and preset searches for common field combinations are ready when you connect your CRM. The searches you trust can then run on their own with Auto Match and Auto Merge. Resources: - [Duplicate Searches](https://dedupe.ly/knowledge-base/search-pads) (KB) - [Rejected Matches](https://dedupe.ly/knowledge-base/rejected-matches) (KB) ### Key Takeaways: Duplicate Search A duplicate search is a saved set of matching criteria (the object, the fields and how strictly to compare each one) that returns duplicate sets from a CRM. Layered searches, one per pattern, sort duplicates by confidence, so clear matches can be merged in bulk while doubtful ones get a closer look. Searches that run on a schedule catch new duplicates as they come in. In Dedupely, Duplicate Searches hold those criteria, with a Match Option for each field and preset searches ready at first login. Rejected Matches keeps false matches out of future results, and Auto Match and Auto Merge handle new matches: Auto Match notifies you when they reach a count you set, and Auto Merge merges them on a schedule. Resources: - [Dedupely product overview](https://dedupe.ly/product) ### Frequently Asked Questions About Duplicate Search #### Do CRMs such as HubSpot or Salesforce detect duplicates automatically? Many CRMs have some built-in duplicate check, but it often looks at only one or a few fields. Depending on the CRM, it might compare email addresses on contacts or domains on companies, list possible duplicates for someone to review, or warn when a new record looks like an existing one. What it checks, which objects it covers and whether it runs on its own differ from one CRM to another, so the CRM's own documentation is the place to confirm. Records that differ in the checked field, such as one person with a work email and a personal email, often get past a built-in check. A duplicate search covers those cases by comparing the fields and strictness you choose, and it can run again as new records come in. In Dedupely, preset Duplicate Searches are ready the first time you log in, and each field gets its own Match Option. Auto Match notifies you when new matches reach a count you set, and Auto Merge merges new duplicates on the schedule you choose. #### Why does a duplicate search return no matches when the CRM clearly has duplicates? Often, the compared fields are blank on many records, or the search is stricter than the data. A search that needs an exact match on last name + email skips every record without an email, and it misses two records where one email has a typo or sits on a different domain. Filters can also narrow a search more than intended. Start by checking that the chosen fields hold data on the records you expect to match. Then loosen one field at a time, for example from exact to similar on names, or to root domain on email and website. In Dedupely, every change to fields, Match Options or filters needs a new Scan before the results update, and you can clone a search to test looser criteria without touching the original. Also check the Excluded tab: Safety Settings move risky matches there, out of the main results. #### How many fields should a duplicate search compare to avoid false matches? A few strong fields together are often enough. A single field, like a name, groups people who just share it, while too many fields make a match so specific that real duplicates slip through when one value is missing or out of date. Choose fields that identify a record on their own or in pairs, such as email, phone, website or an external ID, and set how strictly each one is compared: exact for unique values, looser for names and company names. If a field is unreliable, leave it out instead of loosening everything else. Dedupely's KB recommends combining 2 to 4 key fields in a Duplicate Search. Each field gets its own Match Option, and None leaves an unreliable field out of the match. One search can also hold several match groups to combine different criteria, and Rejected Matches removes a set that isn't a real duplicate from current and future searches. #### When should a duplicate search run automatically instead of by hand? Once its results have proven reliable. Run a new search by hand first and review what it finds over a few rounds. If the sets are consistently real duplicates, such as exact name + email matches from repeat form fills, the search is a good candidate for automatic merging. Searches built on looser or high-stakes criteria, such as similar company names or financial IDs, are better on a schedule that alerts someone instead of merging. Match the schedule to how often records come in: often for busy forms and dialers, less often for occasional imports. In Dedupely, Auto Match and Auto Merge are set per Duplicate Search. Auto Match checks on a schedule and notifies you when new matches reach a count you set; it never merges. Auto Merge merges new duplicates on the schedule you choose, using that search's criteria. Merges in Dedupely can't be undone, so set your Merge Rules first and automate only the searches you trust. #### How do you keep a duplicate search from matching placeholder values like test@test.com? Keep those records out of the search, or leave the field out of the match. Placeholder values, such as "N/A" in a company field, a shared info@ address or a dummy email like test@test.com, are identical across many unrelated records, so an exact search groups them together. A filter that limits the search to records with real values stops the placeholders from pairing up, and comparing on other fields avoids the problem altogether. It's also worth fixing the source, such as a form or integration that fills in a default value. In Dedupely, filters limit which records a Duplicate Search evaluates, and they only change what the search shows, never CRM data. The None Match Option leaves an unreliable field out of the match. Safety Settings, on by default, also hold back matches with poor or example data, such as test@test.com or John Doe, and move them to an Excluded tab, where you can still merge, reject or review them. ### Related Terms and Concepts #### [Matching duplicate records](https://dedupe.ly/glossary/matching-duplicate-records) Matching duplicate records is the stage of deduplication that a duplicate search carries out. Each search is one saved, repeatable set of matching choices, so a team can run the same matching again without rebuilding it. #### [Multi-field matching](https://dedupe.ly/glossary/multi-field-matching) Multi-field matching is how duplicate searches often cut down false matches: they compare several fields at once, such as name + company + phone. The fields a search combines decide how strict it is. #### Rejected and excluded matches Rejected and excluded matches are the sets a duplicate search finds but shouldn't merge. Keeping them out of future results means a search doesn't show the same false matches each time it runs. #### [Auto match](https://dedupe.ly/glossary/auto-match) Auto match runs a duplicate search on a schedule and alerts someone when new matches appear. It's a common next step for a search whose results are useful but still need a person to look at them. #### [AI duplicate matching](https://dedupe.ly/glossary/ai-duplicate-matching) AI duplicate matching finds duplicates across field combinations nobody wrote a search for. It works alongside duplicate searches, which cover the patterns a team knows and wants to automate. #### [Ongoing deduplication](https://dedupe.ly/glossary/ongoing-deduplication) Ongoing deduplication depends on duplicate searches that run again as new records arrive. The searches a team trusts become the ones it schedules, so new duplicates are caught while they're still few. Skip to content