Before You Import That Lead List: A CSV/Excel Checklist for Your CRM

A lead list arrives as a file. A CRM needs records. The gap between those two things is where most teams quietly lose a week of selling — and where some of them also lose the trust of the people they email.

This checklist is the routine we suggest running every time a CSV or Excel file lands in your inbox, whether it came from a one-time pack, a monthly subscription or a custom list. It takes an hour or two the first time and much less afterwards.

1. Open the file before you open your CRM

Download the file, open it in a spreadsheet tool and confirm three things:

• Row count. Does the number of rows match what you expected? A 100-row pack should not contain 100 rows including a header and a blank final line. Count data rows only.
• Header row. Are the column names readable? "Col_3" and "email_address " (with a trailing space) both cause mapping problems later.
• Encoding and delimiter. Accented characters and non-Latin names should display correctly. If the file opened as one long column, it is probably semicolon- or tab-separated rather than comma-separated. Sort that out in the spreadsheet, not in the importer.

While you are in there, delete any column you will never use. Every extra column is another chance to map the wrong field.

2. Write down what each column actually means

This is the step teams skip, and it is the one that matters most. Before mapping, agree on definitions:

• Is "role" a job title, a seniority level, or both?
• Is "company size" employee count or revenue — and in which range bands?
• Is "location" the company's registered address, the person's city, or the region you sell into?
• Are phone numbers one field or several (mobile, switchboard, extension)?

If the file does not answer these questions, ask the supplier before you import. Guessing creates records you cannot segment later, and segmentation is the whole point of buying a curated list.

3. Standardise values in the spreadsheet first

Fix these in the file, not after import:

• Country and region names. Pick one form ("United States", not a mix of US, USA, U.S.A.).
• Job titles. Group obvious variants ("Head of Sales Ops" and "Sales Operations Lead") into a normalised title field, and keep the raw title beside it.
• Dates. Use ISO format (YYYY-MM-DD) for date columns so nothing is interpreted as month-first vs day-first.
• Booleans. Choose yes/no or true/false, not both.

There is a real temptation to hand-edit row by row. Resist it. Use bulk formulas so the transformation is repeatable and you can explain later how a value changed.

4. Deduplicate twice

Duplicates distort everything downstream: two owners calling the same person, inflated counts, and a reply rate that looks worse than your actual work.

Within the file. Use email address (normalised to lowercase, trimmed) as the primary key, then check company domain plus first/last name for the same person under a second address. Spreadsheet tools can flag duplicates; so can most CRMs at import time.

Against the CRM. Export your existing records' email addresses and domain list, and compare. Take a decision on what counts as a duplicate: same person is easy, same company is a judgement call — two contacts at one account may both be useful, and that is a segmentation choice, not an error.

Record what you removed. A short note like "14 duplicates within file, 9 matched existing contacts" saves an argument three weeks later.

5. Add provenance and date columns you will thank yourself for

A lead list is not a static asset; it ages. Before importing, add fields for:

• Source — which pack, subscription month or custom list this came from.
• File delivery date and, if the supplier provides it, verification date.
• What you know about the basis for contact — for example, whether the record was compiled from public business sources, or whether the person has an existing relationship with you.

Why bother? Because when someone asks to be removed, or when a colleague asks where a contact came from, "imported at some point in 2025" is not an answer. Provenance also lets you compare how different lists perform for you over time.

6. Segment before importing, not after

Create the views or list memberships you intend to work from, and assign an owner, before the records land. A simple worked fictional example: a team buys 500 records for a two-rep team. They split the file into three working views — manufacturing in one country, software in another, and a small remainder — and assign 250 each. Nobody imports 500 records and then decides who calls whom.

If your segments turn out to be wildly uneven, that is useful information about your ideal customer profile too.

7. Do a small test batch with a real mapping

Import 10–20 records into a test view or a scratch pipeline stage using the exact field mapping you plan to use. Then check:

• Did the email land in the email field, or in "notes"?
• Did country and industry values create new picklist options?
• Do the source and date fields carry through?
• Does the duplicate rule fire correctly when you re-import the same 10 rows?

Only then run the full import. Testing with 20 rows costs minutes; cleaning 500 wrong records costs a day.

8. Plan the follow-up at the same moment

The import is not finished when the records appear — it is finished when someone knows what happens next. Decide who contacts which segment, on which days, with which message, and where opt-outs get recorded. Keep an explicit rule that a request not to be contacted is honoured across segments and owners, not just in the record where it was received.

And keep expectations honest: a verified email is not a promise of delivery, a reply or a sale. Verification tells you something about an address's format and status at a point in time; it says nothing about whether this person wants to hear from you today. Your bounce handling and reply handling belong in the same process as the import.

A short version for next time

1. Open the file, count rows, fix headers and encoding.
2. Define every column's meaning; ask the supplier where unclear.
3. Normalise countries, titles and dates with formulas.
4. Deduplicate inside the file and against the CRM; log what you removed.
5. Tag every record with source, delivery date and verification date.
6. Segment and assign owners first.
7. Test with 10–20 rows, then import everything.
8. Hand records straight into a follow-up process with opt-outs handled.

If you are describing the file you need in the first place, it helps to specify these fields in your brief — industry, company size, business role and geography are the four we would start with, and the delivery format is CSV or Excel. You can review the current packs, subscriptions and custom-list option on our store page and see which one fits the amount of follow-up your team can realistically handle. If you are still deciding between a file you buy once and a recurring supply, our earlier piece on one-time packs versus a monthly subscription covers that tradeoff in more detail.