About 400 legal professionals found across 32,000 prospects, and one rule that keeps them out of every wave

The problem

The company decided to stop reaching legal professionals. Simple to say. The first reactivation wave had been built before that decision, so it already held nearly 200 of them, some under a segment label and some under plain job titles.

A one time cleanup wouldn't hold. Wave tags rebuild from the database on every run, so a person removed by hand comes straight back. And the pool was set to grow to about 15,000 people.

What we built

One test for "is this person legal", written once and read by everything downstream. It checks three signals: the segment label, the job title, and the company name or email domain. Domains get their own pattern, because a law firm's domain often hides the word inside the stem or ends in .law or .legal.

Each pattern list lives in one place, so adding a term is a one line change. Exception lists keep out the false friends, like a marketing agency with "Legal" in its name. The title check ignores the company exceptions on purpose. A legal lead at a legal tech company is still a legal professional.

Then the rule went into the view every wave and every export reads. Nothing downstream needed to change.

What it produced

• Nearly 200 people came out of the first wave. No record was deleted. Each carries the flag, the reason, and the date.
• 0 rows lost from any other segment.
• About 400 legal professionals identified across about 32,000 prospects. Over 200 of them are stopped before they ever reach a wave.
• 0 flagged people in the export the campaigns read.
• A report only sweep of the remaining wave listed 12 more people for a human to confirm and set aside 14 false friends. The view already holds the 12 out.

What runs now

The rule runs every time the eligibility view is read. As the pool grows, every new person is tested on the way in.

Your scenario

If you've decided a group is off limits, put the rule where the list gets built, not on the list. In Pipedrive, that means a field every filter checks. Our Contact Classifier on Apify reads a job title and returns department and seniority, so a role rule has something clean to test. Send us the group you want kept out and we'll tell you which signals we'd check.

Related tools: Pipedrive, Apify, Contact Classifier on the Apify Store, Contact Classifier

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 August 4, 2026

Last updated August 15, 2026