What the founder believed walking in
A solo personal-injury firm in a smaller market near a big city came to us sure of one thing. The ads cost too much.
The founding attorney had the numbers to back the feeling. A single auto-accident call could cost more than a hundred dollars. A spend target of about $15k a month had been floated, and it felt like a pile of money. So he watched every dollar, and he told us the truth up front: he had been treating this as maybe a short-term experiment he might not be able to afford.
He believed two other things, and both were reasonable. He believed he already owned his website, because he had run it since the mid-2010s and the content was his. And he believed the job was to get more calls, because more leads is what more marketing is supposed to buy.
Stated plainly, his brief was: make the ads cheaper, prove this fast, get me more calls, or I pull out. Nothing about that is wrong. It is what almost every owner says, and it was aimed at the wrong target.
What we saw
We did not argue. We instrumented his account and let his own data talk.
The first number stopped him cold. In one week, 53% of the clicks on his paid ads were competitors clicking to drain his budget. Not prospects. Rivals, burning his money before a real person ever reached him. We paired that number with the dollars we had already saved by filtering it, about $8.2k that week, so he felt the problem and the fix in the same breath. His reaction was two words, and they were not polite.
Then we moved the scoreboard. The auto-accident campaign was generating calls two or three times a day, which looked like winning. But across months it had produced less cases than you can count on one hand. Calls are not clients. You cannot pay your bills with people who saw an ad.
The website was the third reveal. The content was his. The infrastructure was not. He could post through a dashboard, but he could not move the site, change the hosting, or take it with him. Over 500 blog posts sat on shared hosting, set up wrong, working for no one. He had been renting the one thing he most needed to own.
What we decided, and what we chose not to do
Here is the concession we made out loud. The auto-accident channel was lost, and not because the work was bad.
A new competitor had moved in spending more than a hundred thousand a week, pooled money behind a single face, scraping police reports and reselling leads. No single small firm outspends that. So we did not tell him to bid harder on faith. We told him to stop.
The founding attorney got there on his own. If he hired us to build a house, he said, we are not failing because we lack subcontractors. The market quit making lumber and drywall. He absolved the builder, blamed the change in the ground, and stayed.
So the decision was retreat, not double down: exit the channel a better-funded rival had made unwinnable, and redeploy the same money where the math still worked.
What we built
We built four things, in order, each tied to a number he could check.
First, we defended the spend. Behavioral filtering to separate humans from bots, because a genuine lead does not click, click back, then click again. It reads, then acts or leaves. Detection ran around 98%, and we reported the waste we suppressed, about $8.2k a week and about $23,000 in a later month, as value delivered.
Second, we re-wired intake so demand stopped leaking. Call notifications came off the owner’s inbox and went to the front desk. Short and abandoned calls got a warm callback instead of a discard, because a 14-second call is a caller who got impatient, not proof there was no case. Tracking output became a callback list, not a report.
Third, we judged channels by case value divided by cost, not by what a click costs. Listing clicks ran about $12 to $13 against roughly $100 paid-search calls. We modeled a conservative DUI and DWI funnel before spending a dollar: about 8k searches, $15 a click, about 5% click rate, about 20% click-to-lead, about $190 per case. We even downgraded his own 50% sign-up estimate to a modeled 40%, to be conservative, not to aim for the stars. We sized the budget to a small test, not the unusable modeled $6k plan, and used price as a qualifier: “starting at $4,200*” so the firm only paid for prospects already expecting that commitment.
Fourth, we bought the asset instead of renting it. A fully owned website rebuild, priced on the table with no smoke and mirrors, about $50 to $53K, with the design and first year of hosting waived. All 516 posts migrated to dedicated hosting with the web addresses preserved, each page set up properly on a relevance-engine content plan. Your own building, not one suite in someone else’s.
What changed
The defensive results were real, and we reported them as exactly that, not as a promise.
About $8.2k a week and about $23,000 in a later month of wasted spend stopped. The intake leak at the front desk closed. The spend stayed lean and tied to the bottom line, near $4.5k a month rather than the floated $15k, because bidding like a maniac without a reason behind it is not a plan.
We also right-sized the fee. When the retainer could not be carried by personal-injury volume alone, the operator on our side restructured his own pay to a sweat-equity hybrid, about $5,300, on a three-month trial. He named the underperformance first, before the client had to.
The auto-accident channel is designed to stay closed, and the owned website is designed to keep working after the campaigns pause. As the founding attorney was told, even if the marketing stops, the site is worth building. If it produces two good cases over the next four years, it pays for itself.
The lesson
When a business is sure its problem is expensive ads, the answer is almost never cheaper ads.
The complaint is usually three problems wearing one mask. A leak you cannot see. An asset you rent while believing you own it. And a channel you keep feeding because the calls feel like progress, judged by a number that never touches your bank account.
The work is to instrument the truth until the data names the failing part, make the leak visible next to the money already saved, and judge every channel by cases signed over dollars spent. Then buy what you most need to own, and when a channel is genuinely lost, fire it before it fires you.
The edge was never the clever campaign. It was the integrity to stop the leak, own the asset, walk away from the fight you cannot win, and keep building the ground the next win stands on. You already know which of your own numbers you have been afraid to look at.