By Touchbase Technologies · Published

A CRM field-mapping worksheet you can review before importing

A column called Status can mean three different things: whether someone is a customer, whether a salesperson has replied, or whether the person wants marketing emails. If you copy it into whichever CRM field has a similar name, the import can finish successfully while leaving the team with misleading records.

Before you upload a contact file, write down what each field means, where it belongs and what should happen to an existing value. This guide provides a proposed worksheet and fictional examples. It covers preparing and reviewing a mapping; it does not assume that every CRM offers the same import behavior or that Touchbase supports every field described here.

Colleagues meet around an office table with laptops and notes.

Photo by fauxels on Pexels, used under the Pexels License. This is stock photography illustrating a team review, not a photograph of Touchbase customers or an endorsement.

Decide which job the import will perform

Give the import one clear purpose. Are you creating new contacts, updating selected fields on existing contacts, connecting people to companies, or preparing an audience for a campaign? These tasks can overlap, but they require different decisions about identity, relationships and existing data.

Name a person who owns the source file and another who understands the destination fields. They can be the same person in a small business, but both responsibilities need to be covered. Keep the untouched source export in an appropriate private location and use a separate working copy for preparation.

Record the export time and the destination account. A correct mapping applied to the wrong workspace is still the wrong import. Check that the people reviewing the file are authorized to handle its contents, and remove columns that are unnecessary for the intended job.

Write a field contract before matching names

A useful mapping includes more than Source column and Destination field. For each field, describe its meaning and the rule for updating it. Here is a suggested worksheet for a fictional service business. On a small screen, swipe a table sideways to see all its columns:

Source columnMeaningIntended destinationDecision before import
Customer referenceIdentifier in the source systemSupported external identifierConfirm the destination can match on it
EmailCurrent contact addressContact emailReview missing or conflicting identities
Business nameOrganization named by the contactCompany name or associationDo not merge people solely by this value
Request typeReason the person contacted the teamA suitable inquiry propertyDefine the allowed destination values
RelationshipProspect or existing customerRelationship classificationKeep it separate from communication preferences
Marketing choiceRecorded preference for a named purposeSupported preference or exclusion processPreserve the meaning and supporting source
Last customer actionTime of the relevant source eventA clearly defined activity propertyKeep it distinct from import time
Assigned personResponsible source-team memberDestination ownerResolve the actual destination owner identity

These rows are a design example, not guaranteed built-in field names. If the destination has no appropriate place for a field, decide whether to create a supported custom property, keep it in the source system or leave it out. Do not squeeze it into an unrelated field merely to make every column appear mapped.

Confirm how the destination recognizes an existing record

Read the import documentation for the objects you will use. HubSpot's file-format guidance explains supported identifiers for updating existing records and avoiding duplicates. Its options include Record ID, supported unique-value properties and, for contacts, Email. The correct choice depends on the destination object and import process.

A source-system customer reference is not automatically the same thing as a destination CRM Record ID. Label the columns distinctly so a reviewer cannot confuse them. If an import is intended only to update existing records, identify the rows that cannot be matched before approving the upload.

Do not use a person's display name as your only identity rule. Two people can share a name, and one person can appear under different spellings. Likewise, two employees at one company are usually two contacts, not a duplicate simply because the company name matches.

For ambiguous records, use a review queue rather than a confident automatic merge. Keep the original references so the person reviewing the case can trace the decision. A resolved identity is a data-management decision, not proof that the person wants a campaign email.

Define what a blank value means

A blank cell might mean unknown, intentionally cleared, unavailable in this export or simply not relevant to the source. Those meanings are different. Write the expected behavior for each field rather than assuming a blank should erase a value already present in the CRM.

For example, a fictional contact already has an assigned account owner in the destination. The incoming source file has an empty Owner cell because that source does not manage sales assignments. The proposed mapping should leave the destination owner alone unless the business deliberately intends a reassignment.

HubSpot's advanced import instructions describe mapping controls and an option to prevent certain existing property values from being overwritten. Check your chosen product's exact behavior; the presence of an option in one system does not establish that another system behaves identically.

Keep a written decision for changed email addresses, empty preferences, date fields and ownership. These are useful review points because an apparently small overwrite can change who receives a follow-up or who is expected to reply.

Translate categories explicitly

A source may store New, Active and Closed while the destination uses different labels. Build a small translation table and ask the process owner to approve the meaning of each value. Do not infer that New means qualified or that Closed means paying customer.

Fictional source valueProposed interpretationReview decision
New inquiryA request has arrivedMap to the inquiry state, not a purchase outcome
Current customerAn existing customer relationshipKeep separate from marketing eligibility
Waiting on usThe team owes a next actionPreserve an owner and follow-up task where supported
No further contactA recorded contact restrictionPreserve the appropriate exclusion and its scope
Other or blankMeaning is not establishedHold for review rather than guessing

Pay special attention to fields that sound similar but answer different questions. A deal stage describes an opportunity. A relationship describes who the person is to the business. A subscription preference describes a communication choice. One general Status field rarely explains all three clearly.

Test a small set of representative cases

Before importing the full working file, agree on a permitted test using a suitable environment and non-customer data where possible. The purpose is to demonstrate the actual destination behavior, not to send test campaigns to real people.

The following fictional acceptance ledger gives the reviewer something concrete to check:

CaseExpected outcomeWhat the reviewer checks
A completely new eligible contactCreate one intended recordIdentity, mapped fields and correct workspace
An existing matching contactUpdate only approved valuesStable identity and preserved unrelated fields
Two rows with conflicting identity informationFollow the documented review processNo silent merge based on a name alone
A blank source ownerPreserve the approved destination ownerAssignment after processing
A recorded exclusionPreserve the applicable restrictionPreference state before any audience selection
An unsupported categorySurface it for reviewAn actionable error or hold state

Write down the observed result, not just a green import status. If a case fails, stop and correct the mapping before increasing the batch size. A test is useful when it exposes a decision the team has not yet made.

Reconcile the result instead of counting only uploaded rows

After an approved import, distinguish input rows, created records, updated records, skipped rows and errors using the destination's own report. Relationships can also make a row count different from a contact count. Explain the difference rather than forcing the numbers to match an assumption.

For a fictional example, an eight-row working file might represent six distinct people because two people each appear twice with additional source context. That is a reason to inspect the matching and relationship rules, not an instruction to deduplicate by deleting arbitrary rows. Keep a reconciliation note describing the intended people, the observed records and unresolved cases.

Have the next person in the workflow review the result. Can the account owner identify their contacts? Are the intended customer and investor groups separate? Can a campaign owner see the relevant preferences without reconstructing them from notes?

Hand off the mapping with the contacts

Save the approved mapping, test results, export time and reconciliation note together. Name the owner who will maintain the rules when the source changes. A new source value or a renamed field should prompt a review rather than silently inheriting an old interpretation.

For the next stage, use the CRM lead handoff guide to organize who owns the conversation after a contact moves. The signup contact-sync checklist covers a continuing source connection. Those are separate jobs from preparing this one import mapping.

If Touchbase is part of your evaluation, review its current plans and allowances and verify the specific contact fields and CRM connection you need. This worksheet does not promise that a particular integration supports arbitrary custom-field mapping or replaces your destination CRM's import controls.

Questions before approving an import

Can I map fields just because their names look similar?

Check the meaning, allowed values and update rule first. Similar labels can describe different stages, preferences or dates.

Should an import automatically enroll everyone in a campaign?

Treat data arrival and campaign eligibility as separate decisions. Review the intended audience, recorded preferences and sending configuration before enabling a campaign.

What should I keep after the import?

Keep the approved mapping, a suitable source reference, the representative test results and the destination's reconciliation report. That gives the next reviewer a way to understand what changed and why.

Prepared with AI assistance and checked against the sources linked in this guide.