Guide
Moving your sales pipeline from spreadsheets to a CRM
To move your sales pipeline from a spreadsheet to a CRM without losing data: make a copy, keep one row per person or deal, remove duplicates, settle your stages, decide which columns to drop, save each tab as CSV and import 20 test rows before the rest. Last names, leading zeros, colors and notes are what usually get lost. Checked on November 4, 2026.
By the GTM Flow team · Updated · 14 min read
The steps, and what you lose if you skip each one
Moving from a spreadsheet to a CRM takes seven steps: a backup, clean columns, duplicates removed, stages decided, unneeded columns dropped, one CSV per tab and a test import. Almost all the work happens before you open the CRM, because an importer can only read what the file states clearly.
A note on transparency. This guide is written by the GTM Flow team, and GTM Flow is a CRM with its own importer. The steps work with any CRM. What we say about Excel, Google Sheets, Pipedrive and HubSpot comes from their help centers, checked on November 4, 2026; what we say about GTM Flow comes from the product code that same day. Where someone else wins, we say so.
| Step | What you do | What you lose if you skip it |
|---|---|---|
| 1. Backup | Keep the original file and work on a copy | Your way back: removing duplicates in Excel deletes them permanently |
| 2. Columns | One row per record, one value per cell, one header row | Rows that get rejected, or data in the wrong field |
| 3. Duplicates | Normalize, then remove repeats by email, phone or company | Two records for the same person, with split histories |
| 4. Stages | Translate your statuses into a short, fixed list | Deals rejected or placed in the wrong stage |
| 5. Fields | Decide what gets imported, what goes to notes, what gets dropped | A CRM as messy as the spreadsheet |
| 6. CSV | Save each tab separately and check accents and numbers | Whole tabs, plus data Excel had already changed |
| 7. Test | Import 20 rows, check the records, then import the rest | Finding the problem with thousands of records already in |
First, an honest question: do you need to switch? If you handle a few accounts, sell alone and never miss a follow-up, a well-kept sheet can be enough. The move pays off when next steps slip, when several people edit the same list, or when conversations live in email and WhatsApp and the sheet never hears about them. That is our judgment, not a measured figure. If you are sticking with the sheet for now, the sales follow-up process has the formula that flags what is due.
Clean your columns: one row, one value, one header
An importer reads your spreadsheet as a table: the first row names the columns and each row below it is one record. Split contacts and deals into separate tabs, unmerge cells, remove total rows and keep one value per cell. Format phone numbers as text and include the country code.
- One tab per type of data. One for people and companies, another for deals. If a row mixes a person with their deal, repeat it in both tabs with the same company name: that is how they connect later.
- One header row. No titles above it, no subheads, no merged cells. Use recognizable names such as Name, Email, Phone, Company, Stage or Value.
- Remove anything that is not a record. Total rows, separators by month or rep, and loose notes under the table.
- One value per cell. If a cell holds two emails, split them into two rows or choose the right one. Many importers keep one and lose the other.
- First and last name, the way your CRM wants them. Some expect two columns, others one. Check first: this is one of the most common silent losses.
- Phone numbers as text, with country code. Microsoft warns that Excel removes leading zeros and converts large numbers to scientific notation, and that formatting afterward won't restore lost zeros. If you see 3.46E+10, that number can no longer be read: recover it from another source.
- Amounts as plain figures. Remove currency signs, «approx.» and «depends on volume», and use one format for the whole column.
- Turn colors into a column. When saving to a text format, Excel removes all formatting. If red means urgent, add a column that says so in words.
Remove duplicates before you import
Remove duplicates before importing, and on a copy, because in Excel the deletion is permanent. Normalize spaces and capitalization first, then compare by the value that identifies each record: email for people and name for companies. Read the doubtful cases yourself before you delete anything.
- Copy the tab. Microsoft recommends it because, with Remove Duplicates, the duplicate data is permanently deleted.
- Normalize. In Google Sheets, use Data > Data cleanup > Trim whitespace; in Excel, the TRIM function. Lowercase your emails. Sheets notes that non-breaking spaces are not trimmed, and they often arrive when you paste from a website.
- Select only the identifying column. For people, that is email. If you select every column, only rows identical in everything are removed, so two rows for Ana with different notes both stay.
- Look before deleting. Sort by that column and read the repeats. If one row has the phone and another has the title, merge them into one first: removing duplicates doesn't combine anything.
- Fix by eye what no sheet can catch. «Ruiz Productions, Inc.» and «Ruiz Productions Inc» are not equal to Excel. Sort by company and unify the name, or you will end up with two companies in the CRM.
This step cuts the work, but it doesn't replace the CRM's own check. Pipedrive lets you merge duplicates at the final import step, and HubSpot uses unique identifiers such as email or company domain to update records instead of creating them again. If your file lacks that data, the CRM has nothing to compare against.
Decide your stages before uploading anything
A spreadsheet's status column usually mixes stages, tasks and comments. Before importing, translate it into a short list of stages that each begin with a verifiable event, and spell every value the same way across all rows. A task such as «call Tuesday» isn't a stage: it belongs in notes or the next step.
| Your sheet says | What it really is | What you do |
|---|---|---|
| «Interested», «wrote to me» | Stage: interest | Write Interest |
| «Demo on Tuesday» | Qualification: the demo hasn't happened yet | Qualification stage, with the date as next step |
| «Quote sent», «proposal out» | Stage: proposal sent | Use one name |
| «Call», «pending» | A task, not a stage | Pull the task out and find the real stage |
| «No answer» | A lead with no proven interest | Keep as a contact, not a deal |
| «Closed» | Ambiguous: won or lost | Decide row by row, and note the reason if lost |
If your stages aren't settled yet, start from the six in our guide to B2B sales pipeline stages, each with its entry rule. Then decide what to do with history: deals you lost two years ago don't need to come in as open. Leave them out or keep them in the original file, but don't let them inflate your pipeline on day one.
Which fields to drop and which to keep
Not every column deserves to come along. Keep what identifies the person and what you will use to sell this week. Group the useful context in one notes column if your CRM supports it. Calculated values, columns nobody reads and personal data you don't need should stay out of the file.
| Column | What to do | Why |
|---|---|---|
| Name, email, phone, company, title | Import | They identify the person and let the CRM recognize them if you re-upload |
| City, website | Import if you use them | The website helps identify the company |
| Stage, value | Import with deals | They are the pipeline |
| Last activity, comments | Combine into notes | You keep context without a field for everything |
| Commission, weighted value, totals | Leave out | They are calculations: let the CRM or your commission sheet do them |
| Color codes, «ok», «x» | Put into words or leave out | If nobody knows what it means, nobody will next month either |
| ID numbers, birthdays and the like | Leave out unless there's a clear reason | You don't need them to sell B2B |
If you handle data on people in the EU, this step also matters for GDPR. Article 5 (linked in its official Spanish text) requires personal data to be adequate, relevant and limited to what is necessary, accurate, and kept no longer than needed. Switching tools is a good moment to ask why you keep each column. This is general reading, not legal advice.
Don't lose the people who asked you to stop. If your sheet marks them, handle them separately and check how your CRM records opt-outs. A note on a record isn't an opt-out, and the biggest risk in a migration is writing to them again because that column got lost.
Save as CSV and run a test import first
When you save as CSV, Excel writes only the current sheet, so repeat it for every tab. Open the result to check accents and the separator. Then import 20 rows picked on purpose, including the odd cases, open those records one by one, and only then upload the full file.
- One tab, one file. Microsoft reminds you that only the current worksheet is saved and that some features might be lost.
- Check the separator. The default is a comma, but regional settings can change it. A comma inside a company name breaks nothing if the field is in double quotes, as RFC 4180 describes.
- Check the accents. If you see «José», the file is being read with the wrong encoding. Microsoft explains how to open a UTF-8 CSV without breaking them. Regenerate the CSV rather than fixing it by hand if you can.
- Pick 20 difficult rows. A company with a comma, a name with an accent, a foreign phone number, a multi-line note, a value with decimals and a row with no email.
- Import contacts first. That way companies already exist when deals come in, as long as the company name matches across both tabs.
- Open the records and reconcile the totals. Rows in file = created + updated + skipped + rejected. If it doesn't add up, stop and find out why.
- Set a switch date. From that day on, the spreadsheet is read-only. Two sources of truth running in parallel are the most common way to end up starting over.
What each CRM accepts: Pipedrive, HubSpot and GTM Flow
Pipedrive and HubSpot accept XLS, XLSX and CSV directly, with far higher limits than GTM Flow, which reads only CSV and up to 5,000 rows per file. Pipedrive also lets you revert an import within 48 hours. If your file is very large or you want a safety net, they win here.
| Pipedrive | HubSpot | GTM Flow | |
|---|---|---|---|
| Formats | XLS, XLSX and CSV | CSV, XLSX and XLS | CSV only |
| Limit per file | 200,000 rows and 50 MB | 20 MB on free tools; 512 MB and 1,048,576 rows on paid plans | 5,000 rows and 2 MB |
| Tabs | One per import | One sheet per file | CSV: one table per file |
| Duplicates | Offers to merge them at the final step | By unique identifier: Record ID, email or domain, among others | Checked before anything is written, with a preview |
| Undo | Revert within 48 hours, with global admin permission | Not checked for this guide | No undo button |
| Missing fields | Can be added as custom fields | Not checked for this guide | Only the fixed fields in the table below |
Beyond importing, the right CRM depends on how your team sells. Our CRM for small sales teams compares Pipedrive, HubSpot Starter, Close and GTM Flow with their prices.
What GTM Flow imports today, and what it doesn't
GTM Flow imports contacts and deals from a CSV in three steps: upload the file, map the columns and review what will happen to each row before confirming. Data already in the CRM isn't overwritten. It doesn't read XLSX, has no undo and doesn't import custom fields. It's in early access, through a waitlist.
| Topic | What it does today | What it doesn't do |
|---|---|---|
| File | CSV with comma, semicolon, tab or pipe, detected automatically. Reads UTF-8, UTF-16 and Windows-1252 | Doesn't open .xlsx: it asks you to save as CSV first. Maximum 2 MB and 5,000 rows; split anything larger |
| Columns | Suggests a mapping from each column name for you to review. Recognizes common headers from hand-built sheets and from HubSpot, Pipedrive and Salesforce exports | Unmapped columns aren't imported, and custom fields aren't available in the mapping |
| Contacts | Name, email, phone, company, title, city, website and notes. Creates missing companies | There's one name column: a separate Last Name column doesn't come in, so combine both before exporting. With no email, phone or company, the row is rejected |
| Deals | Name, value, stage and company, in the pipeline you choose | No notes, dates or contact: the deal attaches to the company |
| Stages | A stage spelled as in the pipeline is recognized, ignoring accents and case. An empty cell goes to the first stage | An unknown stage rejects the row. There's no screen to create or rename stages: tell us yours when you request access |
| Duplicates | Compares within the file and against the CRM by email, phone, and name with company. The preview gives the reason row by row | Doubtful matches aren't merged: they're skipped for a person to review |
| Re-importing | The same file doesn't duplicate. Existing records only get their empty fields filled | It doesn't overwrite: fixing a value in the file doesn't change the record. An existing deal doesn't update its stage or value |
| Owner | Everything is assigned to the person importing | It doesn't read a rep column. Reassign afterward or import per person |
| Phone numbers | A 9-digit number starting with 6, 7, 8 or 9 is treated as Spanish | Other countries need the country code. An unreadable number is ignored and the row still comes in, with a warning |
| Safety | Either the whole file comes in or nothing does | No undo after you confirm: hence the 20-row test |
Two things we would rather tell you ourselves. If your file has last names in a separate column, including a HubSpot export, combine them first or they're lost. And there's no button to export your contacts and deals, so always keep your original spreadsheet. If you're coming from Pipedrive, you can also sync over its API. We help early-access teams bring their data in.
If you need to import today, or have more than 5,000 rows and don't want to split them, Pipedrive or HubSpot will serve you sooner. If what's missing is email, LinkedIn, WhatsApp and calls on the same record as that data, see GTM Flow, the AI go-to-market platform, how the pipeline works and planned pricing. Access opens in waitlist order.
Frequently asked questions
How do I move my spreadsheet to a CRM?
Make a copy of the file, keep one row per record and one value per cell, remove duplicates, translate your statuses into fixed stages and decide which columns to drop. Save each tab as CSV, import 20 test rows, check those records, and then upload the rest. Start with contacts so companies exist when you import your deals.
Can you import an Excel file directly into a CRM?
It depends on the CRM. Pipedrive and HubSpot accept XLSX, XLS and CSV according to their help centers, checked on November 4, 2026, though each import takes one sheet. GTM Flow doesn't read XLSX, so you save the spreadsheet as CSV first. Even if your CRM takes the original file, clean up columns and duplicates beforehand.
How do I avoid duplicates when importing contacts?
Trim spaces, lowercase emails, remove repeats by email on a copy, and unify company names, since Excel doesn't see «Ruiz Inc» and «Ruiz, Inc.» as equal. Always include an identifier such as email or phone so the CRM can recognize what it already has. Also check how your CRM treats duplicates before you confirm the import.
What happens to my spreadsheet's colors and formatting?
They're lost. Microsoft warns that saving in a text format such as CSV removes all formatting, and that only the current sheet is saved. If a color means something, turn it into a column with words first. Calculated columns such as commissions or totals are usually unnecessary: it's better to let the CRM do those calculations.
Why do zeros disappear or strange symbols appear in my CSV?
Excel removes leading zeros and converts large numbers to scientific notation, so phone numbers should be formatted as text and ideally include the country code. Strange symbols in accented letters are an encoding problem: the file is being read with the wrong one. Regenerate the CSV or open it the way Microsoft's help page describes.
Can you undo a CRM import?
It depends on the CRM. Pipedrive lets you revert one from the import history within 48 hours if you have global admin permissions. GTM Flow has no undo: it shows what will happen to each row first, and either the whole file comes in or nothing does. In any CRM, the safest approach is to keep your original and start with a 20-row test.
The CRM with prospecting built in
GTM Flow brings the CRM, cold email, LinkedIn, WhatsApp and calls into one system, with AI that tells you what to do next. Early access for founders and small sales teams. See how it works
Sources
- Microsoft: Keeping leading zeros and large numbers · checked on
- Microsoft: Find and remove duplicates · checked on
- Microsoft: Save a workbook to text format · checked on
- Microsoft: Import or export text files · checked on
- Microsoft: Opening CSV UTF-8 files correctly in Excel · checked on
- Google: Split text, remove duplicates, or trim whitespace · checked on
- Pipedrive: Importing data into Pipedrive with spreadsheets (updated October 19, 2026) · checked on
- HubSpot: Set up your import file (updated September 11, 2026) · checked on
- RFC 4180: Common Format and MIME Type for CSV Files · checked on
- GDPR, Article 5 (official Spanish text, BOE) · checked on
- GTM Flow · Pricing · checked on
Keep reading
Guide
Sales follow-up process
A sales follow-up process for small B2B teams: when to follow up, which channel, what to log, emails after no response and when to stop. No made-up stats.
14 min11 sourcesGuide
B2B sales pipeline stages
The six B2B sales pipeline stages for a team of 1 to 10: entry and exit rules, what to measure in each, a worked example and a template for your CRM.
11 min4 sourcesComparison
CRM for small sales teams
The best CRM for a sales team of 2 to 10 reps: Pipedrive, HubSpot Starter, Close and GTM Flow, with pricing checked on October 23, 2026.
13 min10 sourcesGuide
CRM for founders
How a founder who sells should choose a CRM: when a spreadsheet is enough, five criteria, which CRM fits each stage and when to switch. Prices checked.
8 min7 sourcesGuide
How much does a CRM cost?
What a CRM costs for 1, 3 and 5 seats: HubSpot, Pipedrive, Zoho, Attio and Close, plus prospecting add-ons, with prices checked on November 4, 2026.
11 min13 sourcesComparison
Pipedrive alternatives
Pipedrive alternatives and competitors: Pipedrive prices checked on 23 October 2026, where it still wins, and 7 CRMs compared channel by channel.
22 min20 sources