Job values vary enormously — a callout and a full installation might come from identical-looking enquiries. Optimising toward enquiry volume drives the cheap ones. The platform needs to learn which enquiries turned into valuable jobs, which means feeding job value back.
Most wasted budget comes from asking one channel to do the other's job. Here is how the work divides for home services.
Nobody opens Instagram planning to buy from you. Meta's job is to interrupt well enough that they want to, then stay in front of them until they do.
Read the detail: Meta Ads for home services
Someone typing your category into Google has already decided they want it. The job is being there, and making the next step obvious enough that they take it.
Read the detail: Google Ads for home services
Why this matters to your bottom line. When both run without shared measurement, they claim the same conversions and you fund the same customer twice. Sorting attribution out first typically recovers more margin than any bidding change, and it costs nothing extra.
Most enquiries are phone calls, so without call tracking the majority of your conversions are invisible. Service area targeting needs to match actual travel willingness. Emergency and planned work are different businesses with different economics.
The number I report on: Cost per booked job, weighted by job value.
Proof of work, response time and trust signals. Before-and-after of completed jobs works well here where it is restricted elsewhere. Local specificity outperforms generic service messaging.
A flat monthly retainer covering both channels, never a percentage of spend. Month to month, and you keep ownership of the accounts and data.
Structure, tracking and 90 days of spend across Meta and Google, checked against your actual margins rather than platform-reported returns.
One source of truth so the two channels stop claiming the same conversions and you stop funding customers twice.
Meta builds demand, Google captures it. Budgets and targets set against what each is actually responsible for.
Money moves toward whatever holds up profitably at higher spend, reported monthly in language you can act on.
You can start with one, and often should. Which one depends on whether people are already searching for what you sell. If they are, Google first. If they do not know they need it yet, Meta first. Running both badly is worse than running one well.
By what each is accountable for, not by a fixed ratio. Typically the channel capturing existing demand gets funded to the point where it stops finding profitable volume, and the rest goes to creating new demand. That split gets reviewed monthly against results, not set once.
Media spend goes directly to Meta and Google on your own billing account — around $2,500 a month across both is where the data becomes readable. My fee is a flat monthly retainer on top, based on account size. Never a percentage of your spend.
I will not quote one before seeing your account, and you should be wary of anyone who does. What I will do on the call is tell you honestly whether your margins, your offer and your budget make paid ads viable at all.
Month to month, 30 days notice, and the ad accounts, pixels and data stay in your name throughout. You keep every bit of campaign history you paid to build, which matters because that history is what the bidding algorithms learn from.
Me. The person on the call is the person in the account, every month. The client list is kept deliberately short for exactly this reason — there is no junior to hand you to.
Bring your accounts and your margins. You will get a straight view of where the money is going and what I would change first — whether or not you hire me.
Book a free strategy call →Each channel covered on its own, plus a way to see where your budget is going today.
How the demand-creation side works for this vertical, and where the money usually leaks.
How the demand-capture side works here, and what stays controllable in a contested auction.
Not sure which channel is letting you down? A written review answers that before you commit.