LATYNEX Labs

GuideEditorial — Our editors' view, not vendor data.

Moving from one CRM to another: a migration checklist for a small team

What to export, how two real CRMs treat your import file, what has to be rebuilt by hand and what to check afterwards. No price, because it depends on your data.

By Nikita LatyshevPublished Updated Prices and specs update from our catalogue

The short answer

A CRM migration is a data project followed by a process project. The data project is exporting records from the old system, mapping them to fields in the new one and importing them in the right order so that contacts, companies and deals stay linked. The process project is everything a file cannot carry: automations, sequences, forms, integrations, email sync, user permissions and reports, which you rebuild and test by hand. There is no fixed price for this, because it depends on how many records you hold, how many custom fields and pipelines you have and how many automations you must recreate; this guide gives you the list to count, not a number.

We use two CRMs whose import rules are published, HubSpot and Pipedrive, to show what a destination expects. Both are read from the vendors' own help pages on 2 November 2026. Whichever way you move, the sequence is the same: inventory, export, prepare, test on a small sample, import in dependency order, rebuild, check and only then switch off the old system. For what each destination costs see HubSpot CRM pricing in euros and which Pipedrive plan a small team needs, and for choosing between them HubSpot vs Pipedrive.

Sources: [1], [2]

The sequence, step by step

A CRM migration in nine steps (editorial checklist, not vendor documentation)
StepWhat to doWhat to check
1 InventoryList objects (contacts, companies, deals, activities, notes, products), custom fields, pipelines and stages, users, automations, forms, integrationsOne count per item: it is your size estimate
2 ExportExport each object with its unique ID and the owner, created date and statusRow counts match the old system
3 CleanRemove duplicates, fix formats, decide which fields to dropA small file you can read in full
4 MapMatch each old field to a destination field or create a custom one firstEvery required field has a source column
5 TestImport 20 to 50 rows of each object into the new systemOpen ten records and compare by eye
6 ImportCompanies and contacts first, then deals, then activities and notesLinks between records survive
7 RebuildRecreate pipelines, automations, forms, integrations and email syncTrigger each one with a test record
8 CheckCounts, samples, owners, dates, amounts, currency and duplicatesA short written sign-off
9 SwitchRun both for a short period, then move users and keep the exportsOld system kept read-only until you are sure

Count before you choose. The size estimate is the sum of your own numbers: records in each object, custom fields, pipelines and stages, automations and integrations. The more of those numbers are large, the more mapping, testing and rebuilding the move needs. We do not put a price or a duration on it, because neither is fixed by the vendors or by us.

Editorial opinion

What two destinations expect from your file

Import rules as the vendors state them (help pages read 2 November 2026)
TopicHubSpotPipedrive
File types.csv, .xlsx or .xls, one sheet; UTF-8 for non-English charactersXLS, XLSX or CSV, one tab, with a header row
Size limits512 MB and 1,048,576 rows a file on paid plans; 20 MB on free toolsUnder 50 MB and no more than 200,000 rows
Required fieldsContacts: Email; companies: Name or Company domain name; deals: Deal name, Pipeline and Deal stagePeople: a name; organizations: a name; deals and leads: a linked person or organization
DuplicatesInclude a unique identifier (Record ID, Email, Company domain name or a unique custom property) or the import creates duplicatesIn the final step you choose how to handle duplicates and can merge instead of creating new records
Linking recordsInclude identifiers for both records in each row; one row per associationOne row can create and link a person, organization and deal
Blank cellsIgnored; they do not clear existing valuesNot stated on the page we read
UndoNot stated on the page we readRevert from import history within 48 hours, by global admins

Sources: [1], [2]

Two points matter most. First, duplicates are a choice you make in the file: HubSpot creates duplicates unless the file carries a unique identifier, while Pipedrive asks you to choose merge or create at the end of the import. Second, links between records come from how you structure the file, so import companies and contacts before deals, and keep the old system's IDs in a column you can search later. HubSpot also lists accepted date, phone, currency and checkbox formats, and says values that break a property's validation rules are not imported, so format the data before you import it. Pipedrive's page lists a 48-hour window to revert an import, which is a reason to import a test batch first and the full file only after you have checked the batch.

Sources: [1], [2]

What a file does not carry

What to rebuild by hand and how to prove it works (editorial)
ItemWhy a file will not carry itHow to test it
Automations and sequencesThe import pages describe records, not workflowsTrigger each with a test record and read the result
Pipelines and stagesThey must exist at the destination before deals import (HubSpot requires a valid pipeline and stage)Import five deals into each stage
Forms and website trackingThey are configuration, not dataSubmit each form and find the record
Integrations and email syncThey connect accounts, not rowsSend and receive a real email on a test record
Users, owners and permissionsOwners are matched by name, email or ID in HubSpot's fileCheck that each record has the right owner
Reports and dashboardsThey depend on fields and stages that now differRebuild one report and compare it with the old one

Editorial opinion

What drives the cost of a CRM migration

Cost drivers, with what you can count (editorial; vendor limits from the import pages read 2 November 2026)
DriverWhy it changes the workWhat to count
RecordsMore rows to clean, map and test; HubSpot allows up to 1,048,576 rows a file and Pipedrive no more than 200,000, so big databases need several filesRecords per object
Custom fields and objectsEach needs a destination field created and mapped before the importCustom fields
Pipelines and stagesThey must exist before deals import; HubSpot requires a valid pipeline and stagePipelines and stages
Data qualityDuplicates and inconsistent formats are fixed before import; HubSpot does not import values that break a property's validation rulesDuplicates in a sample of 200 rows
Automations and sequencesRebuilt by hand and tested; the import pages describe records onlyAutomations
Integrations and email syncReconnected and tested one by oneConnected tools
Users and ownersOwners are matched by name, email or ID in HubSpot's fileUsers
Destination plan and feesSeats are priced per month; HubSpot requires onboarding on Professional and Enterprise, with the euro amount not shown on the page we readSeats and plan
Parallel runningYou pay for both systems while you checkMonths you keep the old one

Sources: [1], [2]

None of these drivers has a fixed price, which is why we give none: the same move is quick for a clean database with one pipeline and slow for one with many custom fields and automations. What cannot be moved automatically is everything outside the record files: automations, sequences, forms, integrations, email sync, permissions and reports, which you rebuild and test. You can do the move yourself with the import tools described above. If you want help scoping or building the website-to-CRM side, LATYNEX Digital, which is operated by the same company that publishes LATYNEX Labs, builds website-to-CRM integrations and automations; its own page on CRM automation cost (opens in a new tab) explains what drives the price, and it has pages on connecting a website to HubSpot (opens in a new tab) and Pipedrive (opens in a new tab). We state no price and no client results for that work, and this guide stands on its own without it.

Editorial opinion

Checks after the move

A post-migration sign-off (editorial; adjust to your data)
CheckHowPass when
Record countsCompare each object in the old and new systemThe difference is explained row by row
Random sampleOpen 20 records per object and compare every fieldNo silent changes to names, dates or amounts
LinksOpen ten companies and see their contacts and dealsEvery link you expect is there
DuplicatesRun the destination's duplicate view or a search by email and domainNo new duplicates created by the import
Owners and stagesCount records by owner and by stageMatches the old system
AutomationsFire each automation with a test recordIt does what the old one did, or you decided it should not
Email and calendarSend, receive and log a messageIt appears on the right record

Editorial opinion

Frequently asked questions

Frequently asked questions

How much does it cost to move from one CRM to another?

There is no fixed figure: the cost depends on how many records, custom fields, pipelines and automations you have and how much of the rebuilding you do yourself. Count those first and ask each vendor what migration help it offers.

How do I avoid duplicate contacts when importing?

In HubSpot, include a unique identifier in the file (Record ID, Email, Company domain name or a unique custom property); without one the import creates duplicates. In Pipedrive, choose how to handle duplicates in the final import step, where you can merge them.

How big a file can I import?

HubSpot allows up to 1,048,576 rows and 512 MB a file on paid plans (20 MB on free tools); Pipedrive's page lists no more than 200,000 rows and under 50 MB. Split larger databases into several files.

Can I undo an import?

Pipedrive's page says an import can be reverted from the import history within 48 hours, by users with global admin permissions. The HubSpot page we read does not describe an undo, so test with a small batch first.

Do automations move with the data?

The import pages describe records only. Plan to rebuild automations, sequences, forms and integrations and to test each one.

Sources: [1], [2]

Sources and checks

Sources

  1. [1]HubSpot knowledge base: set up your import file · cited · checked 2 Nov 2026
  2. [2]Pipedrive help: importing data with spreadsheets · cited · checked 2 Nov 2026

Nikita Latyshev

Founder, LATYNEX Labs

Nikita Latyshev is the founder of LATYNEX Labs, an independent publication about business software and infrastructure. LATYNEX Labs is published by Latynex Trade OÜ, Estonia. How the site checks facts, handles sponsorship and earns money is described on its methodology, editorial policy and affiliate disclosure pages.