Meta ads for insurance brokers: what actually works in 2026
How Facebook and Instagram ads work for independent insurance brokers, what the special ad category changes, and why the broker's own face outperforms everything else.
Most insurance brokers have tried Facebook ads once. It usually went one of two ways: a boosted post that reached a few hundred people and produced nothing, or an agency that ran "lead ads" for three months and delivered a spreadsheet of people who never picked up. Neither is what Meta advertising looks like when it is done properly for insurance.
This is the version that works, and why.
Insurance is a restricted category, and that changes the whole approach
Meta treats insurance as a financial product. Since 2024, ads for financial products in Canada and the US run under a special ad category, which removes most of the demographic and interest targeting that ordinary advertisers rely on. You cannot target by age band, by postcode radius smaller than a certain size, or by the interest signals that made Facebook ads famous.
Brokers hear this and assume the platform no longer works for them. The opposite is true. With targeting stripped back, the creative does the targeting. An ad that opens with "If your home insurance renews in the next sixty days, do not sign it yet" finds the right people on its own, because only those people stop scrolling. The ad set does the reach; the script does the selection.
So the work moves from audience building to writing and filming. That is a better place for it to sit, because a script can be tested and improved every week and a targeting setting cannot.
Your face beats every stock image
Families do not buy life cover from a graphic. They buy it from a person whose explanation made sense and whose face they will recognise on the call. Every campaign we run for a broker is fronted by that broker, filmed on their phone from a script we write and a reference video we send.
The objection is always the same: "I am not a video person." Nobody is. The script is written to be read in short lines, the reference video shows exactly how to frame each shot, and the editor cuts the takes together. The finished ad looks like it cost a great deal more than a phone and twenty minutes.
The structure: test, then blast
A campaign that launches on its full budget from day one is a guess with money behind it. Ours run in two phases.
- Seven days of testing on about 20% of the budget. Two creatives, several headlines, several descriptions and several audiences, run against each other. By day seven the account tells you which combination brings the best leads at the best cost.
- The remaining budget behind the winner. Around 80% of the month's spend, concentrated on the combination that already proved itself.
On a $2,000 monthly budget that is $400 to learn and $1,600 to earn, at roughly $160 a day for ten days. At that spend rate mistakes get amplified, which is exactly why nothing enters the blast phase until the test week has proved it.
What to measure
Not impressions, not reach, not even leads. The number that matters is cost per qualified booked call. A lead who filled in a form and never answered the phone cost you money. A lead who booked a fifteen-minute call and turned up is the product.
Meta reports cost per lead, and it is a useful early signal. On the UK campaigns we publish on our home page it ranges from about £1.50 to just under £7 per form lead, and on a recent Canadian campaign it came in at $12.46, below Meta's own median for similar ad sets. Those numbers tell you the ad is working. Whether the business is working depends on what happens in the ten minutes after the form is submitted, which is the subject of the speed to lead note.
Where the leads go
One more thing that separates a working campaign from a spreadsheet of names. Every lead should land somewhere the broker can act on it in one tap: call, text, note, move along the pipeline. If it lands in an email inbox, it will be called tomorrow, and tomorrow is too late.
That is why our clients run their campaign through an app rather than a report. But whatever tool you use, the principle is the same: the ad is only half the system.
If you are an insurance broker in Canada or the US and want this run for you, the intro call is fifteen minutes and an honest answer.
Book a call