Local-service lead gen on Meta is one of those categories that looks easy on paper and is brutal in practice. Low budgets, impatient clients, a sales team that needs leads yesterday, and zip codes that behave nothing like each other. Here’s the exact playbook I used to drive 5,127 qualified leads on a $37K budget for a US cleaning services brand, with a best-campaign CPL of $2.12.
Rule 1: Geo-target like a local, not a national advertiser
The biggest mistake I see in local-service accounts is treating a multi-city operation like one big campaign. It’s not. Demand, competition, and cost of living vary city-by-city. I split campaigns by metro area (often down to zip code clusters) so I could see which cities were actually profitable and which were quietly draining budget.
Doing this unlocks three things:
- You can reallocate budget to winning metros without rebuilding the whole account
- You can run localized creative (“Same-day cleaning in [City]”) that crushes generic copy
- You can pause underperformers without nuking everything
Rule 2: The offer is the ad
For local services, the offer — not the creative — is doing most of the work. In this account, the winning offer was simple: a first-time customer discount with a clear price and a fast response promise. No gimmicks. I tested three offer variants in parallel:
- Dollar discount (“$40 off your first clean”)
- Percentage discount (“20% off your first clean”)
- Free add-on (“Free deep clean with your first booking”)
The dollar discount won in every single market. Percentages always looked more appealing on paper and always underperformed in practice. I stopped testing that the second time it lost.
Rule 3: Lead form over landing page, almost always
I ran a parallel test of instant lead forms vs. a dedicated landing page. The landing page had better-looking leads but the CPL was 3x higher and volume wasn’t enough to feed the sales team. Instant forms won on math — the slightly lower lead quality was more than offset by the volume, and a tight qualification question filtered out the worst of it.
The form I ended up using had exactly four fields:
- Name
- Phone
- Service area (dropdown)
- When do you need service? (dropdown: this week / this month / just researching)
That last question is the one that really mattered. “Just researching” leads got autoresponded with an email sequence. The other two went straight to the sales team’s phone queue within minutes.
Rule 4: Creative should look like it was filmed on a phone
Every polished brand video I tested lost to rough, phone-shot UGC. A cleaner in uniform, a before/after, a plain voiceover of the offer — that was the template. Production value is actively hurting you on Meta for local services. People scroll past ads that look like ads.
I kept a rotation of 6–8 creatives live at any time, cycling a new one in weekly and killing the bottom performer. That’s it. No sophisticated testing framework, no AI creative optimization. Just ship, measure, cut.
Rule 5: Speed to lead is half the battle
This one isn’t about ads at all — it’s about what happens after. The accounts where CPL kept dropping were the ones where leads were called within 5 minutes of form submission. The accounts where I watched CPL climb were the ones where sales took two hours (or worse, the next day). Meta’s algorithm is smart enough to notice when your conversions actually become customers versus when they ghost. Slow follow-up quietly increases your CPL over time because the algorithm stops trusting your lead signal.
If your client can’t commit to sub-10-minute response times during business hours, tell them to fix that before you touch the ad account.
What $2.12 CPL actually required
The number looks like a flex, but the truth is boring: every single one of the rules above running in sync. Break any one of them and CPL climbs fast. Get all five right and local-service lead gen on Meta starts feeling like a machine. No tricks, no exotic targeting, no AI optimization miracle — just discipline in the boring parts most advertisers skip.