Who this was forA B2B company with inbound site traffic and a separate outbound team.
About two years of website visitor identification data sat in a folder of exports. Most of it had never been contacted. The working assumption was simple: that's a neglected backlog, and outbound should work through it.
Nobody had checked how many distinct people were in there, or how much of it outbound had already touched.
A read only merge with scripts that can be rerun. Each row was classified as a person or a company using the vendor's own rule, with fallbacks. The rule that fired was recorded on every row, and every source file got a checksum.
We excluded 4 of the source files instead of forcing them to fit. The document says which ones.
So it wasn't a backlog. Outbound had been prospecting one population. The website had been attracting another. They barely touched.
That changes what you do with it. Nobody dropped anything. This was an audience outbound had never targeted, and it needs its own messaging.
The merge scripts stay in the client's folder and can be rerun on new exports.
If you're sitting on an untouched pile of records and calling it a backlog, check the overlap before you assign it. Send us the export folder and we'll tell you what's actually in it.
Related tools: Leadfeeder, Clay, Company Identity Resolver, CRM Dedupe Plan
Some of the links on this page are affiliate links.
Client names are always kept confidential at MMG, so they have been removed from this case study.
Published June 23, 2026
Last updated August 13, 2026