Why couldn’t the campaigns use the insurer’s customer data?

For compliance reasons the health insurer could not share customer data, so acquisition campaigns had no way to tell an existing member from a prospect.

How were existing customers identified without that data?

Through a behavioural proxy. Existing customers visit the login page and prospects never do, and that single signal was enough to identify them without one customer record changing hands.

What did the exclusion audience actually do?

It removed existing customers from every lower-funnel acquisition campaign across Google Ads and Meta.

What caused the fall in cost per acquisition?

The page attributes the 60 per cent fall, within two months, to stopping spend on people who could never convert as new customers — not to bidding differently.

Selected work

Advertising · EssenceMediacom

Frank Health Insurance

60% reduction in CPA

Cut CPA by 60% by using behavioral proxies to stop spending acquisition budget on existing customers.

CPA
Analytics

Situation & Strategy

A health insurer was fighting for acquisition in a crowded market, but compliance rules meant they couldn't share customer data with us. Without it, our campaigns couldn't tell an existing customer from a prospect, so we were wasting budget reaching people who were already members.

Tactics

Instead of fighting a losing battle for data, I looked for a proxy. Existing customers do one thing prospects don't: they go to the login page. That single behavioral signal was all we needed to identify them, without touching a single customer record.

Action

I built an exclusion audience using that signal and applied it across Google and Meta, stripping existing customers out of every lower-funnel acquisition campaign.

Results

Cost per acquisition dropped 60% in two months. We didn't change our bidding strategy; we just stopped spending money on people who could never convert as new customers.

This case is written to the same four-part structure as 7 others. Read any one phase across all 8.

View source

Frequently asked questions

Why couldn’t the campaigns use the insurer’s customer data?: For compliance reasons the health insurer could not share customer data, so acquisition campaigns had no way to tell an existing member from a prospect.
How were existing customers identified without that data?: Through a behavioural proxy. Existing customers visit the login page and prospects never do, and that single signal was enough to identify them without one customer record changing hands.
What did the exclusion audience actually do?: It removed existing customers from every lower-funnel acquisition campaign across Google Ads and Meta.
What caused the fall in cost per acquisition?: The page attributes the 60 per cent fall, within two months, to stopping spend on people who could never convert as new customers — not to bidding differently.

More in Advertising

View all