Paid Search
The Smallest Google Ads Account That Actually Works
Most small local accounts are too complicated to manage and too thin to learn. Here is the minimum structure that does the job.
Two failure modes show up over and over in small Google Ads accounts, and they look opposite.
The first is one campaign, one ad group, forty keywords, one generic ad pointing at the homepage. Nothing can be diagnosed because everything is averaged together.
The second is twelve campaigns on a $900 monthly budget, each one starved of data, none of them collecting enough conversions to optimise. Sophisticated structure, no signal.
The fix for both is the same: build the smallest structure that lets you answer the questions you will actually ask.
The questions the structure has to answer
- Which service produced the calls?
- Which geography produced the calls?
- Is branded search inflating our results?
That is it, for most local businesses. If the structure answers those three, it is good enough. Anything beyond that is optimisation you can add when there is enough volume to justify it.
The minimum viable structure
One campaign per service line, not per keyword
If you do residential and commercial work, that is two campaigns, because they have different budgets, different value per job, and different seasonality. If you do one thing, that is one campaign. Do not split "emergency plumbing" and "plumbing repair" into separate campaigns on a small budget — they are the same customer and splitting them halves the learning rate on both.
A separate branded campaign
Searches for your own business name, in their own campaign, with a tiny budget. Two reasons. It is cheap and it stops competitors sitting on your name. And more importantly, it keeps branded traffic out of your performance numbers — branded searches convert at a spectacular rate and will make a mediocre account look excellent if they are mixed in.
Tight ad groups inside the campaign
Five to fifteen closely related keywords per ad group, with ad copy that matches them. If somebody searches "water heater replacement," the ad should say water heater replacement and land on the water heater page. That match is most of what quality score is measuring, and quality score is what determines your cost per click.
Structure exists so you can tell what worked. Every split that does not answer a question is a split that dilutes your data.
Landing pages: the part that gets skipped
Ads pointing at the homepage is the single most common waste in small accounts. The homepage is about the company. The searcher has one specific problem and about five seconds of patience.
- The page headline should repeat the search — literally, in the same words
- The phone number visible without scrolling, on mobile, as a tap-to-call link
- What it costs, or what determines cost. "Call for pricing" loses people
- One form, short. Name, phone, and what they need
This is a website problem as much as an ads problem, and it is usually where the biggest available gain is hiding. Doubling conversion rate on the landing page halves your cost per call without touching a bid.
The settings that quietly waste money
- Search Partners: off, at least at first. It spends on sites that are not Google and converts worse.
- Display Network expansion: off. If it is on inside a search campaign, a large share of your budget is being spent on banner impressions.
- Location targeting: set to "presence" rather than "presence or interest," or you will pay for people in other states who searched about your town once.
- Ad schedule: match the hours a person answers the phone, unless you have after-hours coverage.
- Conversion tracking: configured before the first ad runs, with calls counted and a minimum duration set.
Then leave it alone long enough to learn
The hardest discipline in a small account is patience. Automated bidding needs conversion volume before it is better than manual, and a low-budget account accumulates that slowly. Changing bid strategy every two weeks resets the learning each time.
Set it up properly, let it run for a month, and spend your weekly time on search terms and negatives rather than on restructuring. The account gets better mostly by having waste removed, not by being rebuilt.