I killed HomeAdvisor last month.
Not because the leads dried up. Not because I got mad at their pricing. I killed it because the math stopped working.
For three months, I ran HomeAdvisor against Angi in the same service area with the same budget split. HomeAdvisor was pulling in leads at $38 per job. Angi was at $27. My close rate on HomeAdvisor was 11%. Angi was 16%. That's the difference between a $450-margin lead and a $200-margin lead on the same service call.
So I reallocated the budget. HomeAdvisor's gone. Angi got double the spend.
That's the decision most contractors never make. They add platforms, they keep them running, and they wonder three years later why they're drowning in expense. The real growth move isn't finding more leads. It's ruthlessly optimizing which sources deserve your money.
Know Your Breakeven Number
Before you can kill anything, you need to know when something's broken.
Your breakeven is simple: the cost-per-lead you can afford to pay and still make money after the job is done. Here's how I calculate mine.
I charge $150 for a service call. My average HVAC call runs $3,200 in revenue. My cost to dispatch, fuel, parts, and labor is about 55% of that, so I net $1,440 per job. If I'm paying $60 per lead and closing 1 in 8, that's $480 in lead cost per closed job. That still leaves me $960 in margin.
But that's my absolute ceiling. Realistically, I want lead cost to be no more than 15% of gross revenue per job. On a $3,200 call, that means I can't spend more than $480 total to acquire it. If I'm closing 1 in 7, my max cost-per-lead is about $68.
Write down your numbers. Don't estimate. Pull your last 90 days of invoices and calculate your actual average revenue per closed job. That's your anchor.
Track the Ratio, Not Just the CPL
Cost-per-lead is useless without close rate.
I can have cheap leads all day long and go broke. My Thumbtack account was pulling in leads at $22 per. Looked incredible. Except my close rate was 6%. That made each closed job cost me $367 in lead spend.
Meanwhile, my Angi leads were $27 each with a 16% close rate. Same job cost: $169 in lead spend. The Angi "expensive" leads were actually three times cheaper.
This is where most contractors miss the signal. They see the CPL number and make a decision. They should be looking at the inverse: revenue per lead spent, or what I call your "acquisition efficiency ratio."
Here's the formula:
Average job value × Close rate ÷ Cost per lead = Acquisition efficiency ratio
I watch this number more than I watch the CPL. If it's trending down for a platform, that's my cue to dig in and figure out why.
Run a 90-Day Head-to-Head
You can't make a good kill decision on gut feel or six months of data.
Here's what I do: I pick two platforms I think are competitors (or one I want to test and one I know works). I give them equal budget for 90 days. Same service areas. Same seasons. I track every lead source, every quote, every close.
After 90 days, I have real data. Not vibes. Not "I think Thumbtack's better." Numbers.
When I ran that HomeAdvisor vs. Angi test, I didn't kill HomeAdvisor after week two. I ran the full test. Turns out HomeAdvisor's cost-per-lead actually got cheaper in month three as their algorithm learned my preferences. But it never caught up to Angi's close rate, and that's what mattered.
The Kill Decision
Once you have the data, the decision is mechanical.
Rank your platforms by acquisition efficiency ratio. Cut everything below your breakeven. If a platform is hitting but inefficiently, cut it and reallocate that budget to your best performer. Let the best performer run harder.
Here's what mine looked like in Q1:
| Platform | CPL | Close Rate | Acq. Eff. Ratio | Decision |
|---|---|---|---|---|
| Angi | $27 | 16% | $169 | Keep, increase budget |
| Thumbtack | $22 | 6% | $367 | Kill |
| HomeAdvisor | $38 | 11% | $345 | Kill |
| Google Local Services | $31 | 14% | $221 | Keep, monitor |
| Referrals (tracked) | $0 | 28% | $0 | Reinvest savings into lead follow-up |
Thumbtack and HomeAdvisor got cut. Angi got the freed-up budget. Google Local Services stayed because it was still profitable, even if it was third.
The hardest part isn't the math. It's the ego. You spent time setting up that account. You got to know your rep. Killing it feels like failure.
It's not. It's discipline. You're not failing. You're optimizing.
What Happens After You Kill Something
When I killed HomeAdvisor, I didn't just pull the plug and forget it.
First, I watched my volume. Angi can only absorb so much budget before the cost-per-lead climbs. I needed to make sure killing HomeAdvisor didn't create a lead gap in month two.
It didn't. Angi handled the extra budget smoothly.
Second, I told my team. "We're going all-in on Angi for the next quarter. Stop worrying about HomeAdvisor leads, because there won't be any." Clear direction, clean handoff.
Third, I watched the freed-up budget. That money didn't vanish. I either reinvested it in Angi, or I used it to fund something new I wanted to test (like Google Local Services).
The mistake contractors make is killing something and leaving a hole. You kill a source, you need a plan for where the volume's coming from instead.
The Quarterly Audit
I run this same analysis every quarter.
Not because I'm obsessed (okay, maybe a little). But because platforms change. Their algorithms shift. Your close rate drifts. A platform that was your worst performer in Q1 might be your best in Q3 if the algorithm learned your preferences.
I've brought back platforms I killed before. But I've also killed the same platform twice because it kept underperforming.
The quarterly audit isn't punishment. It's maintenance. Just like you'd check your HVAC system before summer hits, you check your lead sources before the season changes.
Do the math. Run the test. Make the call. Your margin will thank you.