Data quality
CSV Enrichment Checklist: Prep Before Upload
Run this before you enrich a CSV of leads: the columns every row needs, phone, email and address clean-up, deduplication, file limits and checks afterwards.
By the GTM Scoop team · Published · 4 min read
Most enrichment problems are file problems. A phone column full of extensions, two emails pasted into one cell, ZIP codes that lost their leading zero in a spreadsheet: each one costs matches you would otherwise have had.
How to enrich a CSV of leads walks through the upload step by step. This is the checklist to run before you upload, and again when the results come back.
The checklist at a glance
- Decide what one row is.
- Make sure every row can be matched.
- Clean the phone column.
- Clean the email column.
- Split the address.
- Remove duplicates before you spend.
- Check the file against the limits.
- Map the columns and read the coverage.
- Check the results.
1. Decide what one row is
- One row per person or one row per company. Mixing the two makes the results hard to read.
- A single header row, with one column per field.
- Your own ID column, such as a CRM record ID or a row number, so you can join the results back to the source.
- A source column (the event, the partner, the form) if the file mixes several origins, so you can see later which source matched best.
Keep the company name, website and job title columns too. They are not what the match runs on, but they make the results much easier to check by eye, and a rep reading the enriched file will want them.
2. Make sure every row can be matched
GTM Scoop matches each row on phone number, then email, then address and ZIP. For company data, the mapping needs a phone or an email column and an address column with ZIP.
- Filter your sheet for rows with no phone, no email and no address. They cannot match, so fix them or set them aside before you start.
- Fill gaps from your own systems first. A phone number from the CRM or a billing address from finance costs nothing, and every row that cannot match still counts in the credit check.
3. Clean the phone column
- One number per cell. Move extensions and second numbers into columns of their own.
- Use one format for the whole column, and keep to US numbers: GTM Scoop's coverage is US only.
- Remove placeholders such as 000-000-0000 or n/a rather than leaving them to fail.
- When a row has several numbers, map the main company line or a direct line, not a fax number.
4. Clean the email column
- One address per cell, trimmed, with no names or angle brackets around it.
- If you hold both personal and work addresses, keep them in separate columns and map the work one.
- Team inboxes such as info@ or sales@ reach a shared mailbox rather than a person. Keep them if they are all you have, but expect them to say less about who you are dealing with.
- Fix obvious typos in domains, such as a letter missing from a well-known provider's name.
5. Split the address
- Street, city, state and ZIP each in a column of their own.
- Five-digit ZIP codes, stored as text. Spreadsheets drop leading zeros, so a Boston ZIP such as 02110 can turn into 2110.
- Two-letter state codes or full state names, used the same way all the way down the column.
- The street address rather than a PO box where you have both: it says more about where the company actually is.
6. Remove duplicates before you spend
Each matched row uses a credit, so a person who appears three times can cost three credits. Deduplicate with the same keys the matcher uses:
- Start with identical email addresses.
- Then rows with the same phone number.
- Then rows with the same street address and ZIP, checking by eye where the names differ, since several people can share one office.
- Keep the most complete row of each group, and note which IDs you merged.
7. Check the file against the limits
- Save it as a CSV.
- An upload holds up to 10,000 rows. Split bigger files.
- Give each file a name you have not used before: GTM Scoop refuses a second upload with the same file name.
- Check your credits. A start needs at least as many remaining enrichment credits as the file has rows (Starter includes 200 a month, Growth 500), though only the rows that match are charged.
8. Map the columns and read the coverage
In Connect Data, upload the file and check each suggested mapping: an email column mapped as a name will not match anything. Before you commit, you see how many rows can be matched. If coverage is lower than you expected, the cause is usually a missing or mis-mapped phone, email or address column, so fix the mapping before you enrich.
9. Check the results
- Compare the number of matched rows with what the coverage check showed. A big gap is worth understanding before you run the next file.
- Look at which signal each row matched on: phone, email or address.
- Spot-check a sample against a second source before a large send: open the company's website, or compare with a recent email thread.
- For rows that did not match, add what is missing and run them again. Often it is one column, such as a ZIP or a phone number, that the source never captured.
- Download the results as a CSV, or keep them in GTM Scoop as a saved list.
Keep a copy of the cleaned file next to the results. The next time the same source sends a list, you will know which fixes it needs, and the checklist takes minutes instead of an afternoon.
When the list is ready, send it to your CRM or outreach tools. Records get a Domain column when they have a website or a work email to take it from. That makes de-duplication in the CRM easier.
Try it in GTM Scoop
Upload your file, check the suggested mapping and see how many rows match before you spend a credit.