Google Ads for home services: not every suburb is worth the same bid

Home services businesses live and die on two numbers nobody in their ad account is tracking: how far the van has to drive, and what the average job is worth once it gets there. Until the campaign structure reflects those two numbers, bidding is guesswork dressed up as automation.

The situation

A trades business covering a wide metropolitan area with one search campaign, one budget and a radius target around the depot. Calls were coming in. The owner's complaint was not volume — it was that the team seemed busier than ever and the month's numbers had not moved.

Which is exactly what happens when the ad account buys a lot of small, distant jobs. Every one of them is technically a conversion. Collectively they fill the diary with travel time.

What the audit actually found

• One campaign, one radius. No way to bid differently for a suburb five minutes away versus fifty.
• All calls counted equally. A thirty-second price enquiry and a booked installation were the same conversion.
• No job value data in the account. The business knew average job value by service type; the ad account did not.
• Service types mixed together. Emergency repairs, routine maintenance and installations sharing one ad group, so the cheapest clicks dominated the budget.
• No ad scheduling. Spending overnight and on public holidays when nobody could answer the phone.
• Missing negatives. "DIY", "how to fix", "parts", "second hand", "jobs" and "apprenticeship" all pulling clicks.
• Generic landing page. Every ad landing on the homepage, with no suburb or service relevance and the phone number below the fold on mobile.

The most expensive lead in home services is a cheap one an hour away. It looks like a win in the ad account and a loss on the day sheet.

The approach, in order

1. Get real job data before touching the account

   Average job value, average travel time and win rate — by service type and by suburb group, from the invoicing system. This takes an afternoon with the office manager and it determines everything else. Without it, every geographic bid decision is a preference rather than a decision.

   Usually a clear picture emerges immediately: two or three suburb clusters produce most of the profitable work, a middle band is roughly break-even, and an outer band has been quietly subsidised for years.

2. Then segment geography into value tiers

   Suburbs get grouped into tiers — core, secondary and marginal — based on job value net of travel. Each tier gets its own campaign or its own location bid adjustments, and the marginal tier gets a decision: bid down hard, restrict to high-value service types only, or exclude entirely.

   Location targeting also gets set to "presence" rather than "presence or interest", which alone removes a surprising volume of out-of-area clicks.

3. Then split by service type

   Emergency, maintenance and installation are three different businesses sharing a phone number. Different urgency, different margins, different ad copy, different times of day. Once separated, each gets a budget matched to what it is actually worth, and the emergency campaign can bid aggressively at the hours it converts without funding low-margin clicks at 2am.

4. Then make calls count properly

   Call tracking with a minimum duration threshold so brief price checks stop registering as leads, call recordings reviewed weekly at first, and — where the business will do it — booked-job values imported back into the account as offline conversions. That last step turns the whole thing from cost-per-call optimisation into value optimisation, which is the difference between a busy diary and a profitable one.

   Call extensions, location extensions and a click-to-call button visible at the top of the mobile landing page. In this vertical the phone is the conversion; everything else is a fallback.

5. Then landing pages per service and area

   A page per core service, with suburb coverage listed and genuine local proof — jobs completed nearby, licence and insurance details, response time commitments. The phone number stays visible while scrolling on mobile. Ads point to the page matching their service, not to the homepage.

6. Then report by suburb and service

   The monthly review becomes a table the owner recognises: spend, calls, booked jobs and revenue per suburb tier and per service. That is the report that lets a business decide to open a second depot, or to stop servicing an area — decisions worth far more than a percentage point of click-through rate.

What changed

Two things move at different speeds here. Cutting marginal geography and adding negatives has an immediate effect — wasted spend stops in the first week, and the same budget starts buying calls closer to the depot. Average job value moves more slowly, over roughly two to three months, as the mix shifts toward installations and away from small distant repairs.

Total call volume normally falls. If the owner is used to judging the account by how often the phone rings, that has to be discussed before anything is switched off — the target is jobs completed and revenue per van day, not calls received.

What I would do differently

I would put call recordings in front of the owner in week one instead of describing them. Ten minutes of listening to actual enquiries does more to settle the "which leads are worth having" question than any spreadsheet, and it makes every later decision faster.

Common questions