Skip across the chapter
Quota Playbook

Contact imports: 3 match fields to check first

Quota Playbook Editorial Team · Published · 13 min read

Before importing contacts, check organization name, phone number, and email against your CRM's matching rules.

Key takeaways

  • Confirm which fields your CRM compares.
  • Standardize organization names and phone formats.
  • Check shared lines, generic inboxes, and email typos.
  • Test a small sample before importing the full file.

A rep opens the pipeline, spots a prospect they remember, and finds three contacts, two deals, and a note from someone who left the company last spring. Much of that mess starts at import time. A spreadsheet of leads lands in the system, and the system decides which rows are new and which rows already exist.

This post covers the three fields in Pipedrive's person-matching example, and what to check in each one before you click import. The examples come from Pipedrive's documentation, because it explains its matching rules in plain terms. Your CRM may work differently, so read its help pages for the exact rules. The habits still carry over: know which fields the system compares, clean those fields first, and decide what happens when a match is found.

Why import duplicates hurt a deal review

Duplicates are more than untidy. A contact that exists twice can carry two separate activity histories, so the rep you ask about last week's call may have no record of it. An account split across two organization records can hide how many open deals a company really has. In a deal review, a manager may approve a forecast based on a pipeline that looks bigger, or more fragmented, than it really is.

The answer is not to stop importing. Events, partner referrals, and marketing campaigns can supply import lists. The answer is to set up each import so the CRM can tell which rows describe the same person or company, and then to inspect proposed matches if the CRM offers a preview.

The three fields Pipedrive checks

When Pipedrive recognizes a person in an import, it checks for a matching organization name, person phone, and person email. Its guide uses a sample file with several rows named "Jon Snow" to show how this works. One of those rows is checked against the organization name, phone, and email of the existing record, and the guide contrasts it with a later row that finds no match on those fields for the existing "Jon Snow". ( Pipedrive's import guide)

The point for you as a manager is simple. Two rows can share a name and still be different people. The fields the system compares are the real test, so the quality of those fields is what you need to control before import.

Field 1: Organization name

Organization name is one field in Pipedrive's person-matching example. Check it alongside the other identifiers. It is not unique, though. Pipedrive's guide says you can have organizations with the same name as long as you always specify the address field. ( Pipedrive's import guide) In practice, two different companies called "Summit Logistics" can both be valid, and a list that leaves out addresses gives the system no way to tell them apart.

Check the organization column in three ways before import.

First, look for variations of the same name. "Summit Logistics", "Summit Logistics Inc.", and "SUMMIT LOGISTICS" may be one company or three. Pick one format, such as the legal name without punctuation, and apply it to the whole file. Do not rely on reps to fix this by hand, because they will each fix it differently.

Second, fill in the address column for every organization row you can. If the address is blank for some rows and filled for others, the matching behavior can change from row to row, and you will not notice until a duplicate shows up in the pipeline.

Third, look for rows with no organization at all. A contact with no organization gives the matcher less to work with. Decide before import whether those rows should be loaded, held back, or attached to an account by someone who knows the company.

A useful rule for managers: if the organization name is the only thing you know about a row, treat that row as unverified. It may be a real new account, or it may be a second copy of one that already lives in the CRM under a slightly different spelling.

Field 2: Phone number

Phone numbers look like a clean identifier. Digits are digits, and two rows with the same number seem obviously to be the same person. The trouble is formatting. The same line can be written with brackets, with dashes, with spaces, or with a country code in front, and to a person those are the same number. Check your CRM's documentation to learn whether formatting affects phone matching. A shared office or front desk line also links several people at one company to a single number.

Pipedrive's guide includes an example in which a row's phone number ends up mapped onto the same phone number created by an earlier row. The quoted line reads: "Instead, it maps row 2's unique phone number onto the same phone number created in row 1." ( Pipedrive's import guide) The lesson is that the output can link rows in ways you did not intend. Read the phone column the way the CRM reads it, not the way you read it in a spreadsheet.

Before import, do three things with the phone column.

Normalize the format. Choose one pattern, with or without a country code, and apply it to every row. Strip spaces, brackets, and dashes if your CRM stores plain digits. A consistent format makes an exact match possible.

Sort the column and scan for repeated numbers. Possible explanations include the same person listed twice, a shared line for several people, or a data entry error. Each needs a different decision. Duplicate rows for the same person can be merged. Shared lines should be stored once, with a note, and not used as the only way to reach a decision-maker. Entry errors should be fixed at the source.

Flag shared lines. A front desk number should not become the phone record for a senior buyer. If you cannot tell whether a number belongs to one person, leave it blank for that row, and let the rep add the direct line after the first call. A blank phone is easier to fix later than a wrong one that has already linked two people.

Field 3: Email address

Email can help identify a contact. An address is normally tied to one person at one company. Even so, email can mislead in two ways. People change jobs and keep old addresses in your files. And generic inboxes, such as an info or sales address, can appear on many rows, which makes them a poor identifier for any one person.

Person names can repeat. Pipedrive's guide notes that there is always a chance you will have contacts with the same name, and gives "John Smith" as the example. ( Pipedrive's import guide) A matcher that relies on name alone can combine different people who share a name. Check the name together with organization, phone, or email to help distinguish contacts.

Before import, check the email column for three things.

Blank cells. A row with no email loses one of its strongest matching points. If you have a phone number or organization instead, check those identifiers and make sure someone has confirmed the row.

Generic inboxes. Addresses such as info@, sales@, or hello@ belong to a company, not a person. Decide whether they should become contacts at all. If they should not, remove them. If they must be kept, store them in a way that does not make them look like a personal record.

Typos. Check for missing domains, doubled dots, and swapped characters. An email typo can prevent an email match, although another matching field may still connect the records. Run a simple validation pass before import, and fix the obvious ones.

How the matcher combines the three fields

This is where many managers get surprised. Pipedrive's guide says that if the system finds that any of those three fields matches another person in the database or the import file, it will merge them together and add any new data included in the import. ( Pipedrive's import guide) That is an "any match" rule, not an "all fields must match" rule. A single shared phone line or a single reused email can pull two people into one record.

This is useful when you want duplicates caught. It is risky when the file is messy. A row that matches an existing record on one field can merge with it, adding new data. Before you import, ask a simple question: if this row matched an existing record on just one field, would I be comfortable with the merge? If the answer is no for a group of rows, fix those rows first.

Deals follow different logic, and the difference matters. Pipedrive's guide says that if you import more than one deal with the same name, those deals will not be considered duplicates, and all of them will be imported. ( Pipedrive's import guide) A deal name is not a safe key. If your deal list has repeated names for renewals, different product lines, or different sites, the system will keep them all. That is correct in some cases and a source of noise in others, so decide the rule before the file goes in.

A preparation pass for a deal review list

Consider a manager who has a long file of leads from a software conference. The sales lead wants all of them in the CRM before the deal review, and the manager has little time to prepare. Here is how the three-field check would run in practice.

The first pass is the organization column. The manager sorts it alphabetically and sees "Harbor Systems", "Harbor Systems Inc", and "Harbor Sys." Those are probably the same company. The manager asks which legal name the company uses, fixes the variants to match, and then checks the address column. Most rows have a street address. The rows that are blank get filled in from the company website, or get marked for review by the rep who met that contact at the booth.

The second pass is the phone column. Most numbers come in as plain digits, but some have a country code and others have brackets and dashes. The manager normalizes all of them to a single pattern. Then the manager sorts by phone and finds one number that appears on several rows, each with a different job title. That is almost certainly a shared switchboard. The manager keeps one row with that number, removes it from the others, and asks the reps to get direct lines during their follow-up calls.

The third pass is email. A couple of rows have no address at all. One is a reply from a person who only gave a name and a phone number, so it goes on the hold list. Another has a sales inbox, which the manager replaces only after confirming that the named person and email belong to the original lead. A few rows have obvious domain typos, and the manager fixes them. Nothing about this work is complicated. It is just careful, and it is done before the file touches the CRM.

The last step is a test import. The manager loads a small sample into a test area, inspects proposed matches if the CRM documents a preview, and checks the sample results before importing the full file. The manager investigates any unexplained merges. When the deal review starts, the pipeline shows real companies with real contacts, and nobody has to ask why a prospect has two histories.

Illustrative example: Check a small sample

Take 10 rows from a lead file. Sort by phone, then check organization and email alongside each repeated number. Hold any row whose identity you cannot confirm. Test the remaining rows and compare the resulting records with the file before proceeding.

A pre-import checklist for managers

Use this list before any file goes into the CRM. Most of these steps take little time, and skipping them tends to cost far more after the deal review has already been scheduled.

  1. Name one owner for the final file. Several people can send changes, but one person should decide what goes in.
  2. Confirm which fields your CRM compares when it recognizes a person or an organization. Write those fields on the file's cover note so everyone sees them.
  3. Normalize organization names. Choose one format and fix the obvious variants.
  4. Fill in addresses for every organization row you can, and mark the rows where you cannot.
  5. Normalize phone numbers to one pattern, and flag shared lines.
  6. Clean email addresses. Remove blanks you cannot fill, decide on generic inboxes, and fix typos.
  7. Hold back rows that contain only a name. Enrich them or leave them out.
  8. Run a test import on a small sample, and inspect proposed matches if your CRM supports a preview before you run the full file.
  9. Use the duplicate-recognition option. Pipedrive's guide says that whenever you run an import, you are given the option to recognize and merge duplicates together. ( Pipedrive's import guide)
  10. Reconcile the counts. Compare the rows in the file with the records created and merged, and explain any gap before the file is closed.

Clean up duplicates already in the CRM

Some duplicates are already in the system before the new file arrives. If your concern is contacts that exist in the account today, Pipedrive's guide points to its Merge Duplicates feature for that job. ( Pipedrive's import guide) Run that cleanup before the deal review, not during it. A pipeline cleaned up in advance is far easier to trust than one you have to explain on the call. For that separate task, see clean duplicate contacts.

When you merge two records, choose the primary one first, then compare the fields that differ. The newer record may have a better phone number or a corrected title. The older one may hold the activity history and the deal links. Do not assume either record is correct by default. Keep the one with the most useful history, and copy over only the fields you have verified.

Make it a team habit

Tools only solve part of the problem. The rest is a short standard that reps and managers follow every time. Write down the three match fields, the format you require for each, and who signs off on an import. Put imports on the calendar well before the deal review, so there is time to inspect the results and fix what looks wrong.

Managers set the tone here. If a rep loads a list straight into the CRM the night before a review, that is a process problem, not a character flaw. Give people a clear path: a standard file, a named reviewer, and a deadline. Revisit the standard now and then, especially after any incident where a duplicate caused a bad forecast call. Carry that check into your weekly pipeline review.

Your next step

Today, take a copy of your next import file and sort its phone column. Flag repeated numbers for confirmation before loading it.

Related guides: Check 7 deal fields before your next coaching review · Check 5 task fields before the next coaching 1:1.

Questions before import

Does every CRM use these match fields?

Check your CRM's documentation for its exact matching rules. This post uses Pipedrive's person-matching example.

Can I inspect individual matches before saving?

Use a match preview only if your CRM documents that feature. Otherwise, test a small sample and inspect the resulting records.

What should I do with uncertain rows?

Hold uncertain rows back for confirmation. Ask the rep who knows the contact to resolve the identity before import.

Sources

  1. Update team roles in 3 checks after the setup change
  2. Lost seller phone recovery: 4 steps for 2FA lockouts
  3. HubSpot quote checks: 5 limits to run before email

More plays