A respected brokerage with almost no digital pipeline.
The client had been in business for over twenty years. Strong referral base, two locations, a small but loyal walk-in flow. They were not in trouble — they were in stasis. Every year looked like the one before.
Their previous Google Ads account had been managed by a generalist agency on a percent-of-spend model. Broad-match keywords, near-zero negative-keyword work, no conversion tracking beyond “click on phone number”. The account ran $2,400/month for eighteen months and produced maybe forty bound policies. The math wasn't close to working.
What we inherited at kickoff:
The state of the account
One campaign. Three ad groups. 247 broad-match keywords, half of which were irrelevant (“insurance jobs vancouver”, “icbc claim form”, “life insurance for dogs”). No negative keyword list. No conversion tracking inside Google Ads itself — just a phone number swap that fed everything into one bucket. Quality Score average: 3.4 / 10.
The owner's diagnosis was correct: “We're paying for a lot of clicks that aren't even our customers.”
We're paying for a lot of clicks that aren't even our customers.
The first month was almost entirely demolition. We turned off every broad-match keyword, replaced them with phrase and exact match variants, built a 600-word negative-keyword list from search-term reports, and rebuilt conversion tracking from scratch — calls, forms, and call-duration thresholds (a 12-second call is rarely an opportunity).
Then we waited. Soft-launched the new build with a capped daily budget at 60% of the old spend. The first thirty days were quiet on purpose. We were learning what the account actually looked like when the broad-match noise was gone.
