Case study · 90-day engagement
An Orlando pool service company was spending $4,400 a month on Google Ads. The CPL had crept up from $38 to $71 over two years. The agency running it kept saying "Google is more competitive now." We pulled the search term report and what we found explained $1,800 a month of the spend in a single afternoon.
The challenge
The owner had been frustrated for a year. CPL kept climbing. Booked jobs were flat. The agency dashboard showed steady "engagement metrics" and the recommended response was always to spend more.
When we ran a 12-month search term report:
The agency had not run a search-term report in 11 months according to the change history log. Their internal automation had been running on autopilot. The owner was paying $890 a month for "campaign management" that was effectively no management.
The plan
Pulled 12 months of search terms, classified each one for service relevance, bulk-added 287 negatives. About half were obvious (table, noodle, exercises, tournament). The other half were judgment calls (specific competitor brand names, geographic queries outside the service area, price-only shoppers). Spend dropped 38% the same week and lead volume actually went up because the budget was no longer being wasted.
Restructured 4 broad ad groups into 11 service-specific ad groups: weekly maintenance, equipment repair, green-pool cleanup, salt cell replacement, pump replacement, heater repair, leak detection, vacuum repair, tile cleaning, opening, closing. Each ad group has its own keywords, ad copy, and ideally a matching landing section.
Moved 80% of keywords from broad-match (the previous default) to phrase + exact match. Broad-match has its place when you have huge budgets and tight conversion goals; for a $4k/mo budget on a small service business, it bleeds money.
The website had one services page. We split it into 11 anchor-linked sections so each ad group could deep-link to the relevant section. Quality score climbed across the board.
Set up a recurring Monday-morning search term review. Bad terms get negated weekly. New negatives average 8-12 per week. The waste does not creep back in.
AI tools we used
We do not pretend the work happens by hand. Three of our internal tools are the reason this engagement moved as fast as it did. Each of them replaces what used to be days of human time.
Bulk-classifies 12 months of search terms into "service-relevant," "tangential," or "kill it." Turned what would have been a 3-day audit into a 4-hour task.
Takes one bad term and proposes the 8-12 related variants you should also negate. Caught 87 negatives we would have missed.
Runs every Monday at 6 AM, surfaces new bad search terms from the past week, drafts the negative keyword list for review. Kills the "we forgot to look" problem.
The execution
| Week | What we shipped | What moved |
|---|---|---|
| Week 1 | 287 negatives bulk-added, audit complete | Spend down 38% the same week |
| Week 2-3 | 4 broad ad groups rebuilt into 11 service-specific groups | Bid algorithm has clean signals |
| Week 3 | Match-type discipline: broad → phrase + exact | CPL drops below $50 for first time in 8 months |
| Week 4-6 | Service-section landing page split, anchor links | Quality score moves from 5/10 to 8/10 average |
| Week 8 | Weekly search term hygiene cron deployed | Waste prevention, not just cleanup |
| Week 12 | CPL stable at $34, 2.1× more leads on same budget | Owner takes the savings as profit |
The numbers
The $1,800/mo waste broke down to:
The previous agency was charging $890 a month for management. Their automation had been running on autopilot since the day the account was set up. The owner had been told that this was normal. None of it was normal. A weekly 30-minute look at search terms is not a feature. It is the bare minimum.
Asked my old agency for a year why my CPL kept going up. They blamed Google. Mark sent me one screenshot of the search terms after the second week and I understood the entire problem.
Keep reading
Thirty minutes. We pull up your account on screen. We tell you what we would do, what we would not do, and whether we are the right fit for your business.