Help Center

Docs

Importing Data & CRM Migration

Bring your records, your history and your money over from the system you are leaving

What a move is

One run that reads everything you upload, tells you what it found, and writes nothing until you say so.

A move starts by asking where your records are coming from, then takes every file you have in one go. We read them, show you what is in them, ask about the few things we cannot settle on our own, and only write once you have seen what would happen. When it is done you get a report of every row, and a way to put it all back.

Open the Admin Panel from your account menu, then pick Import & Migration under Data in the admin sidebar. The address bar reads /admin/import once you are there, and Start a move begins one.

Who can import
Import & Migration sits inside the Admin Panel, and what opens it is the Bulk Import permission on your role. That permission is enough on its own: it is the one admin page a role without admin panel access can still reach, so you do not need to be an admin to run a move. A role with admin panel access reaches the page too. Bulk Import is a permission rather than a job title, and it carries a scope: some roles hold it for the whole workspace, some for their team, and some only for the records they own. If you cannot reach Import & Migration, ask whoever manages roles in your workspace to give your role Bulk Import.

The steps

In this order, and you can leave and come back to any of them.

1

Files

Drop everything you have at once, from as many files as it takes. Nothing is imported yet. We work out what each file is, and say so on the file itself, so a file we have read the wrong way can be corrected before anything else happens.

2

What we found

Every column we read, and what we are doing with it, in three words: Matched, Best guess, Not matched. There is no score and no percentage behind those words. Change any of them here, and a column we have no home for can be given one or left out.

3

Decisions

The few things nobody but you can settle: two records that might be one, people whose company is not in your workspace yet, a date that could be read two ways. Each one is a question with the real rows behind it. You can leave the rest for later without holding the move up.

4

Money

If your files carry invoices, payments, credits or refunds, this step checks that what we are about to write adds up to what your old system said, and shows you the comparison before it writes anything.

5

Preview

What would happen to each kind of record, counted, with a sample you can read. Nothing is written while we check. This is the last screen before the move runs.

6

Moving

We write the records, in the published order below, and show you where we are as it goes. You can close the tab. Rows that could not be written are collected as they happen rather than sinking the run.

7

Report

What came across, what did not and why, per record type and per row, kept for good. The report is also where the undo lives.

What we can bring over

We will bring these in this order so the links hold.

You do not have to bring all of it. Pick the tabs you want on the first screen and the rest waits. What you do pick arrives in the order below, so a record that points at another one always finds it.

Contacts and companies

The people you work with, the companies they work for, and the history against them.

Suppressions, Unsubscribes, Companies, People, Addresses, Company tags, Person tags, Contact roles, Relationships, Contact groups, Group members, Email preferences, Consent records, Activities.

Sales

Deals, the stages they move through, and the things you sell.

Pipelines, Pipeline stages, Product categories, Products, Prices, Variants, Bundles, Deals, Deal tags, Deal contacts, Deal line items, Quotes.

Money

Invoices, payments, credit notes, recurring billing and tax rates.

Tax codes, Exchange rates, Vendors, Inventory locations, Opening stock balances, Invoices, Credit notes, Payments, Refunds, Recurring schedules, Sales orders, Purchase orders.

Support

Tickets and their conversations, queues, surveys and the knowledge base.

Queues, Queue members, Knowledge base categories, Surveys, Tickets, Ticket conversations, Knowledge base articles, Survey responses.

Projects and work

Projects, tasks, the templates behind them and the time booked to them.

Project templates, Task templates, Projects, Project tags, Tasks, Task tags, Task comments, Task dependencies, Time entries.

Scheduling

Booking pages, working hours, resources and the bookings already made.

Booking pages, Booking page hosts, Working hours, Resources, Bookings.

Templates and settings

The words and choices your workspace runs on, brought over before the records that use them.

Invitations, Teams, Team members, Tags, Custom fields, Lifecycle stages, Lead statuses, Territories, Email preference categories, Email templates, Snippets, Signatures, Email labels, Picklist options.

Files and photos

Attachments that belong to a record. This one comes in a later step.

What goes in each column

The headings we read, what lands where, and an example of the shape we expect.

Your headings do not have to match ours. We read the spellings below and a good many more, and anything we are unsure about comes back to you on the second step rather than being guessed at. Three record types are shown here because most moves start with them; every other record type publishes the same list on its own starter file.

Companies

FieldDescription
company*Company name. For example Acme Manufacturing. Also read as company name, account name, organization. The name we match everything else against, so keep it spelled the way your team says it.
domainWeb address. For example acme.example. Helps us spot the same company arriving twice under two spellings.
phoneMain phone number. For example +1 415 555 0142. Also read as phone number, main phone.
emailMain email address. For example hello@acme.example. Also read as email address.
typeWhat they are to you. For example customer. Also read as company type, account type. A word we do not recognise stops the row, so the accepted words are listed with this record type.
statusHow active they are. For example active. Also read as account status.
industryIndustry. For example Manufacturing.
cityCity. For example Oakland.
countryCountry. For example United States.
assigned_toWho owns it. For example dana@yourcompany.example. Also read as owner, account owner. Use the work email of somebody already in your workspace, or leave it blank.

People

FieldDescription
first_name*First name. For example Dana. Also read as first name, given name. A single full name column works too. We split it for you.
last_nameLast name. For example Whitfield. Also read as last name, surname.
company*Company name. For example Acme Manufacturing. Also read as company name, account name. A contact needs a company. Bring the companies file first and these link themselves.
emailEmail address. For example dana@acme.example. Also read as email address, work email.
phonePhone number. For example +1 415 555 0188. Also read as phone number.
titleJob title. For example Operations Lead. Also read as job title, position.
lifecycle_stageWhere they are in your funnel. For example customer. Also read as lifecycle stage. We keep your own words where your workspace already uses them.
linkedin_urlProfile link. For example https://www.example.com/in/dana. Also read as linkedin.
assigned_toWho owns it. For example dana@yourcompany.example. Also read as owner.

Deals

FieldDescription
name*Deal name. For example Acme line upgrade. Also read as deal name, opportunity name.
company*Company name. For example Acme Manufacturing. Also read as company name, account name. A deal needs a company. Bring the companies file first.
stageStage. For example proposal. Also read as deal stage. We keep your own stage names where your workspace already has them.
amountValue. For example 48000. Also read as value, deal value.
currencyCurrency. For example USD. Give this whenever your deals are not all in one currency.
expected_close_dateExpected close. For example 2025-09-30. Also read as close date, expected close date. A date we cannot read stops the row rather than landing as a blank.
assigned_toWho owns it. For example dana@yourcompany.example. Also read as owner.

The words we recognise

Status columns take words, not codes, and these are the ones we already know.

Activities, the type column

call, email, meeting, note, task, follow up, followup, incoming email, outgoing email, logged call, logged email, logged meeting, logged note, phone call, send letter, send quote, lunch, deadline, mail, phone, event, in person, online meeting, video call, demo, presentation, sms, text, chat, letter.

Activities, the priority column

low, medium, high, urgent.

Deals, the stage column

discovery, prospecting, research, qualification, qualified, qualifying, proposal, proposal price quote, quote, negotiation, negotiation review, contract, contract sent, closed won, won, closed lost, lost, appointmentscheduled, qualifiedtobuy, presentationscheduled, decisionmakerboughtin, contractsent, closedwon, closedlost, lead in, contact made, needs defined, proposal made, negotiations started, perception analysis, id. decision makers, value proposition, needs analysis.

A stage word we do not know lands the deal in Discovery, and we say so on the row.

People, the lifecycle stage column

subscriber, lead, mql, marketing qualified lead, marketingqualifiedlead, sql, sales qualified lead, salesqualifiedlead, opportunity, customer, evangelist.

Tasks, the priority column

low, lowest, medium, normal, mid, high, highest, urgent, critical, blocker.

Tasks, the status column

todo, to do, in progress, working, started, review, in review, awaiting review, done, completed, complete, closed, finished, not started, deferred, waiting on someone else, open.

Tickets, the status column

open, new, reopened, in progress, working, waiting on customer, pending, awaiting reply, waiting on internal, escalated, resolved, solved, fixed, closed, done, cancelled, canceled.

Your own words are kept where your workspace already uses them. A word we do not know is raised on the Decisions step rather than dropped.

Two records that look like one

Never decided silently. You choose the answer and you choose what we match on.

When a row and a record already in your workspace look like the same thing, the Decisions step shows you both and offers three answers: One person, Two people, Decide later. Nothing is merged or skipped until you pick one.

You also choose what we compare on. The options are Company name, Website domain, Their old system's id, Email address, and the one you pick is used for the whole run so the answer is the same on every row.

When the people in a file name a company you have not brought over, the same step offers Create the companies too, Bring the people in without a company or Leave them out. Bringing your companies over first is still the tidier route, because a company you already have keeps everything else you know about it.

Support tickets that arrive in two parts

A ticket list without the conversations is half of what you think it is, and we say so first.

This looks like a ticket export. It has your tickets but not the messages inside them. We can bring the tickets in now so your history is searchable today, and add the messages when you upload the full export.

The two answers are Bring in the tickets now and Wait, I will get the messages first. Taking the tickets now costs you nothing later: when the fuller export arrives, we add the messages to the tickets that are already here and create no second copy of any of them.

Your tickets are here without the messages inside them. Upload the full export when you have it and we will add the messages to the tickets that are already here.

What is not coming across.

Said before the move runs, not discovered afterwards.

Every move shows a block under this heading for the system you said you are leaving. It lists what that system holds that we do not bring across, in whole sentences, on the screen where you decide whether to go ahead. It is also repeated on the report, so the answer to what happened to something is on the same page as what happened to everything else.

When we hold nothing recorded against a system, the block says so rather than going quiet: We have nothing recorded for this one yet. Check your export against the list of columns we read and tell us what is missing.

Some of it is made here instead, and the block says which: Saved replies are set up in Laureo rather than carried across.

Files and sizes

What we read, and how much of it at a time.

We read CSV files, Excel workbooks saved as .xlsx, and JSON. Older .xls workbooks are not read, so save those as .xlsx or CSV first.

  • One file can hold up to 10,000 rows.
  • One file can be up to 10 megabytes.
  • You can undo a move for 30 days after it finishes.
  • You can undo a single file import for 24 hours after it finishes.
  • A CSV or Excel file needs its column names in the first row. Commas separate the columns, and text is read as UTF-8.
  • One cap covers every record type. If a file is over it, split it and upload the pieces into the same move.
  • A file you upload but never import is kept for 7 days and then deleted, along with its rows. Start the upload again if you come back after that.

We take this on for moves of up to 50,000 records. Above that we will talk it through with you first.

Staged Files

A file put in place ahead of an import, and the id it gets back.

A staged file is a spreadsheet that has been put in place for an import but has not been imported yet. Staging checks the file against the requirements above, reads its header row, counts its rows and gives the file an id. Nothing is created or changed in your CRM by staging alone. The import wizard does not work this way: it reads your file in the browser and carries it straight through column matching into the import in one sitting. Staging is the other way in, and today it is what happens to a file you attach to a chat message.

The id is how the later steps address the file, so you map columns and run the import against the staged copy instead of uploading the same spreadsheet again. Reading a staged file back answers with the file name, the row count and the column names only, never the rows themselves, and only the person who staged the file or a workspace administrator can read it. A staged file you never import is kept for 7 days and then deleted along with its rows, so start the upload again if you come back after that.

Laureo can stage a file for you. Attach a CSV or JSON file to a chat message and ask what is in it: the attachment is staged for import, and you get back the column names, the row count and, if you name the kind of record, suggested column matches and the columns that record type needs. Staging from chat imports nothing, so you still choose the column matches and confirm the import yourself. It needs the same import rights the wizard does.

Putting it back

Every finished move can be undone from its own report.

The report page carries the undo. It deletes what the move created and returns what the move changed to the state it was in beforehand, and it asks you to type UNDO first, because anything you or your team changed by hand since the move will go with it. The window is on the list above.

The two undos are not the same permission. A single file import can be put back by anyone who can import, for 24 hours after that file finishes. A whole move is put back by an administrator, under the Admin permission on their role, for 30 days after the move finishes.

When something is not right

The handful of things worth checking first.

  • We named the wrong system. The first screen is a question, not a guess we make for you, and you can change your answer at any point before the move runs. When a file argues with your answer, we show you what in the file made us think so.
  • A column went somewhere you did not expect. Everything we read is on the second step with the word we used for it. Set it yourself there, or leave it out. Nothing is written until the Preview step.
  • Dates read the wrong way round. A file whose dates could be read two ways is raised as a question rather than settled for you, and your answer is applied to the whole file.
  • People whose company is not here yet. Bring your companies over first, or answer Create the companies too on the Decisions step and we will create them from what the people file says.
  • Rows that could not be written. They are listed on the report with the reason on each one, and you can download just those rows, fix them and bring them into the same move.
  • The move seems to have stopped. It has not. A move runs on our side, so you can close the tab; the report shows where it got to whenever you come back.