When skip tracing results or property lists pass through Microsoft Excel on the way to a CRM, a few problems come up again and again: leading zeros disappear from IDs and ZIP codes, long IDs lose their final digits, and phone numbers or dates change format. Any of these can make an import update the wrong record or create duplicates. This checklist covers Microsoft Excel only.
Digits become numbers. Excel treats a column of digits as a number by default, so 00412 can become 412.
Long IDs lose precision. Microsoft lists Excel's number precision as 15 digits (Excel specifications and limits). An ID longer than 15 digits that Excel reads as a number can have its final digits replaced with zeros.
Phone numbers lose their format. A number stored as +12025550101 can be read as a plain number, dropping the plus sign.
Dates get reformatted. A 2026-10-15 value can be converted to a date and saved back in your computer's regional format.
Quotes do not protect values. Wrapping a value in quotation marks inside the CSV does not stop Excel from converting it when you open the file by double-clicking.
[ ] Import instead of double-clicking. Microsoft's guide (Import or export text files) describes Data > Get & Transform Data > From Text/CSV, as well as the older Import Text Wizard, which lets you set a column to Text to keep leading zeros.
[ ] Stop automatic type detection. In the From Text/CSV preview, choose not to detect data types, or select Transform Data and set the columns yourself.
[ ] Set ID, ZIP, phone and date columns to Text during import. Changing the format afterwards does not bring back digits that were already dropped.
[ ] Treat any ID longer than 15 digits as Text, every time.
[ ] Keep one row per contact in the import file, with numbered phone, status and date groups side by side (see the DNC data dictionary).
[ ] Do not sort or reorder phone columns on their own. If numbers need to move, rebuild the import file from your research rows so each status stays with its number.
[ ] Keep a contact-wide opt-out in its own column. A custom opt-out column may not change your CRM's built-in subscription or opt-out setting, so update that setting as well.
[ ] Spot-check three long IDs and three leading-zero values against the original file.
[ ] Save as CSV, then reopen the saved file in a plain text editor (not Excel) and confirm IDs, ZIP codes, phone numbers and dates are unchanged.
[ ] Do not double-click the saved CSV to "check it" in Excel and save again, since that can reintroduce the same changes.
Digits lost from a long ID: Excel cannot recover them. Go back to the original export or source file.
Leading zeros lost from fixed-length values such as five-digit ZIP codes: a formula such as =TEXT(A2,"00000") can restore them, but only when every value has the same known length. Do not use this on IDs of varying length.
Unsure what changed: compare the edited file against the original in a text editor or a file comparison tool before importing anything.
[ ] Choose the import mode on purpose. According to HubSpot's guide (Import records for a single object), a row without a Record ID can create a new record, while an update-only import skips rows that do not match an existing record.
[ ] Map Record ID when updating exported records. The same guide explains that Record ID takes priority over other unique identifiers, such as email, when matching records.
[ ] Import a small preview first, export those records back, and compare them field by field with your source file.
Other CRMs handle matching differently, so check your own CRM's import documentation for how it identifies existing records.
Copy the checklist into your project notes or list management SOP.
Run through it each time a file passes through Excel, not only the first time.
Test the long-ID and date behavior once with a sample file in your own version of Excel, since behavior can vary by version and settings.
For the complete import process, see the Medium guide How to Keep DNC Status Attached to the Right Phone Number.
Sources checked 2026-10-21. Maintained by Locate Next (locatenext.com).