Google Ads is genuinely good at what it does. Somebody types that they want the thing you sell, and you can be the first result within an hour of deciding to be. Nothing else in marketing works that directly.
The catch is that the platform's defaults are designed for advertisers with enough budget to spend their way through a learning period. Set up an account by clicking accept on everything Google suggests and you will spend real money finding out what does not work. In a market the size of ours, where a monthly ad budget is often a few thousand dollars rather than a few hundred thousand, there is no room for that.
These are the three mistakes that cost the most, in the order they do damage.
Mistake 1: running ads without conversion tracking
This is the big one, and it is astonishingly common. The account runs, the budget goes out, clicks come in, and there is no measurement of what happened after the click.
Two things go wrong at once.
You cannot tell what is working. You will know that a campaign got four hundred clicks. You will not know whether those clicks produced twelve enquiries or none, or which search terms produced them. Every optimisation decision after that is guesswork dressed up as analysis.
Google cannot help you. Modern bidding is automated, and the automation optimises toward whatever you tell it to count. With no conversion data, the best it can do is chase clicks. And it will get you clicks, cheaply, from people who were never going to buy. The system is working exactly as instructed. It was instructed badly.
Setting it up properly means counting every way an enquiry actually arrives:
- Form submissions, tracked on the thank-you state rather than the button click, so you count completions rather than attempts.
- Phone calls. If most of your business comes by phone, and for a lot of local businesses it does, an account that only counts forms is blind to most of its own results. Google can track calls from ads and calls from a number swapped onto your site.
- WhatsApp clicks. This one gets missed constantly here. If your main call to action is a WhatsApp button, the click on that button is your conversion, and it needs to be tracked as one.
Then check the quality of what you counted. This part matters. Not every form fill is an enquiry. If your conversion count and your actual enquiry count disagree, trust the enquiry count and go fix the tracking. A conversion number nobody believes is worse than none, because it gets optimised toward.
Mistake 2: sending every click to the homepage
Somebody searches for a specific thing. Your ad says you do that thing. They click, land on your homepage, and now have to find it themselves.
Most will not. They will look for a few seconds, not immediately see what they came for, and go back to the results, where they click a competitor and you have paid for the privilege.
The rule is simple: the page should be about the thing they searched for. If the ad is for one service, the landing page is that service, with the relevant detail and an obvious way to get in touch. Not a homepage carousel, not an "about us" narrative, not a services page with twelve entries where theirs is fifth.
Google notices too. Landing page relevance feeds Quality Score, and better Quality Score means you pay less for the same position. A relevant page is cheaper to advertise as well as more likely to convert. The effect compounds in both directions.
What that page needs is unglamorous:
- The thing they searched for, named in the heading, in their words
- Enough detail to make the decision: what it involves, what it costs or depends on
- Proof: reviews, work you have done, how long you have been at it
- One obvious action, repeated, and reachable without scrolling on a phone
- Speed, on mobile data
You do not need a bespoke landing page for every keyword. You need the page to be about the thing. If you already have a good service page, use it.
Mistake 3: letting the defaults spend your money
The account creation flow is designed to get you live quickly, and along the way it opts you into several things that quietly waste budget. Most of the money we see wasted in local accounts is here.
Broad match with no negative keywords. Broad match lets Google decide what is close enough to your keyword. With no conversion data to steer it and no negatives to constrain it, "close enough" gets loose. You end up paying for people looking for jobs, free options, DIY instructions, and the thing next to the thing you sell. Start with tighter match types, watch the actual search terms report weekly for the first month, and build the negative list from what you see. This is the single highest-return hour in the account.
Location targeting set to presence or interest. This is the default, and it means your ads can show to people merely showing interest in your area rather than being in it. In a small market with a large diaspora, that difference is significant. You can spend a meaningful share of budget reaching people who are not here and cannot buy. Set targeting to people in your locations. Then check the report of where your traffic actually came from, because the setting alone does not always behave the way you expect.
The Display Network switched on inside a Search campaign. Search campaigns offer to extend onto Display, and it is on by default. Display clicks are cheap and mostly worthless in this context, and because they are cheap and numerous they swamp your averages and make the campaign look better than it is. Turn it off. Run Display separately if you want it at all.
Auto-applied recommendations. Google will offer to apply its own suggestions automatically, including adding keywords and changing bidding. Some suggestions are fine. Some will broaden your targeting in ways you would never have chosen. Turn auto-apply off and review recommendations yourself.
Performance Max with nothing excluded. Pmax will happily spend a large share of budget on people already searching your business name, traffic you would have got free, and report it as excellent performance. If you run it, exclude your brand terms so you can see what it is actually contributing, and keep an ordinary Search campaign running alongside for comparison.
The setup that works on a small budget
If you are starting or restarting:
- Get conversion tracking in first. Forms, calls, WhatsApp. Do not launch without it.
- One campaign, tightly targeted, in the locations you actually serve, presence only.
- A small set of keywords you are certain about, on tighter match types.
- A negative keyword list from day one, and check the search terms report weekly at the start.
- Send each ad group to the page about that thing.
- Manual or click-based bidding until you have conversion data, then move to conversion-based bidding once there is something to learn from.
- Leave it alone between reviews. Daily fiddling prevents the system from ever settling.
Small and disciplined beats broad and hopeful. You can widen once you know what converts.
The question to ask before any of this
Ads amplify whatever is already happening. If your site is slow, your enquiries go unanswered for a day, or your Google listing has the wrong hours, ads will buy you more traffic into the same leak, at a cost.
So before the budget goes on: is the page you would send people to actually good? Does somebody answer the phone? Is your Google Business Profile in order, given that it earns you the same customers for free? And do you know what an enquiry is currently worth to you?
When those are in place, ads are one of the most direct tools available. When they are not, ads are an expensive way to discover the problem.
If you would rather this was set up and managed by people who do it daily, that is a service we run, and the first thing we will do is check whether you should be spending at all.