Should a small business run Google Ads at all

That is the whole decision, and most agencies will not lead with it because the second and third conditions are not things they get paid to fix.

Paid search is the only channel where you can spend a thousand dollars in a week and have nothing to show for it by Friday. That is not a flaw. It is the point. Money moves fast in both directions, which means a good setup pays back quickly and a bad one drains quietly until you decide ads do not work for your business.

They probably do work. The page probably does not.

What do Google Ads actually do for a small business?

They buy you position on a search you did not earn yet. Nothing more.

That is worth stating plainly because the promise gets inflated. Ads do not build reputation. They do not make you memorable. They put your name in front of somebody who is already typing a phrase that means they want to buy something, at a moment when they are willing to click.

The value comes from intent. Somebody searching "emergency plumber San Jose" at 9pm is not browsing. They have a problem, and they will call one of the first few options they see. You are renting a spot in that moment.

The cost comes from the same place. Everyone else wants that moment too, so the price per click reflects it. LocaliQ, which publishes annual search advertising benchmarks under the WordStream name, puts the 2026 average cost per click at $8.33 for Home and Home Improvement, $8.00 for Dentists, $4.62 for Beauty and Personal Care, $2.05 for Restaurants and Food, and $5.42 across all industries. Those are medians rather than means, LocaliQ sells advertising management, and the 2026 edition does not publish its sample size. The 2025 edition of the same series disclosed 16,446 US campaigns with a minimum of 64 campaigns per category. Treat the figures as vendor research and a starting estimate, not as your price.

What you pay per click is not set by your bid alone. Google's own documentation describes Quality Score as a diagnostic estimate of how relevant your keyword, your ad and your landing page are to the search, reported on a one to ten scale using expected click-through rate, ad relevance and landing page experience. Relevance is priced. The page is not just where the click lands. It is part of what the click costs.

How do I know if my page can carry paid traffic?

Look at what your existing traffic already does. If you get organic and referral visitors now and almost none of them contact you, paid visitors will behave the same way, except you will be paying for each one.

This is the piece that gets skipped. An owner sees flat lead volume, concludes the problem is not enough traffic, and buys traffic. The traffic was never the constraint. The page was.

A page that converts poorly does not fail quietly when you add ads. It fails at the speed of your daily budget.

You cannot test your way out of this either. Run the arithmetic. At 500 monthly visits and a 3% conversion rate you get 15 leads a month (derived). Split across two versions of a page, that is seven or eight leads each, and a two-lead difference is inside ordinary month to month noise.

Kohavi, Deng, Longbotham and Xu, writing in the KDD 2014 paper "Seven Rules of Thumb for Web Site Experimenters," report that at Bing most changes fail and the ones that succeed move key metrics by 0.1% to 1.0%. That is web-scale search data from Microsoft and LinkedIn, not local data, and the domain gap matters. What transfers is that real effects are small and small samples cannot see them. At local volume you fix the page with judgment, not with a test.

The most common version of this problem is sending ad clicks to the homepage. A homepage is built for people who already know who you are, and it does a different job than a page built to convert a stranger. The two are not interchangeable, and paid traffic exposes the difference immediately.

For anything visual, proof does more work than copy. If you sell a service where the result is something people look at, the photos on the page change whether a stranger asks for a quote more than any headline you write.

What do the numbers look like before I spend anything?

Five lines, on paper, before a card goes into the account.

Line one, price the clicks. Take LocaliQ's category figure as a placeholder until you have your own. Home services, $8.33.

Line two, buy the clicks. A $2,000 monthly budget divided by $8.33 is 240 clicks (derived).

Line three, turn clicks into leads. At LocaliQ's 8.05% conversion rate for the category, 240 clicks produce 19 leads (derived) at a cost per lead of $105 derived. That is roughly consistent with the published $90.92 cost per lead for the category, which is a useful sanity check on your own arithmetic.

Line four, turn leads into customers. This is the line owners skip. If you close a third of the quotes you give, 19 leads is 6 customers (derived) and your cost per customer is $333 derived.

Line five, compare to gross margin, not to revenue. If your average job bills $1,200 and your gross margin on it is 35%, the job contributes $420. At $333 to acquire, you are ahead by $87 per customer before overhead. Thin, but real.

Now rerun line three at a 4% conversion rate, which is what a page that buries the phone number will often produce. Same 240 clicks, 9 leads (derived), 3 customers (derived), $667 per customer (derived). The same campaign that cleared its margin now loses $247 on every job.

Nothing changed in the ads account between those two versions. The only variable was the page.

Where does this reasoning break down?

In three places worth naming, because a rule you cannot break is a rule you will apply wrongly.

When demand is already yours. If most of the search volume for your name is people who already know you, ads on that traffic will look spectacular and prove nothing. A brand campaign converts well because those people were coming anyway. That distortion is real and it is why defending your brand name is a different campaign with different arithmetic than going after new demand.

When the phone, not the page, is the destination. For a lot of trades the conversion is a call, and the website is a formality. Google's own Local Services documentation is explicit that responsiveness to customer inquiries feeds ad rank, and that missed calls may negatively affect it.

When the honest answer is that nobody knows. No published study with a disclosed sample measures what share of small businesses profit from paid search. That is a real gap and I would rather say so than fill it. What can be measured is your account, which is why the arithmetic above is built entirely out of numbers you already own.

Are search ads even the right kind of ad?

For a lot of service businesses, no. Google's Local Services Ads sit above the regular search results, charge per lead instead of per click, and require a verification process that a lot of competitors will not bother completing. The structure, the eligibility, and the tradeoffs are different enough that Local Services Ads deserve their own decision before you assume search ads are the default.

There is also a management cost, and each fee structure pushes the manager toward different behavior. Understand that before you sign anything, because what you pay for management changes what your manager optimizes for.

Sterling Sky's 2026 field data, drawn from Jepto and Places Scout tracking, records local pack ads rising from around 1% of tracked mobile reports in early 2025 to around 22% by December 2025, and Local Services Ads rising from around 11% to 31% of tracked queries in the same window. That is an agency reporting on its own tracked keyword set, not a census, so read it as direction. The direction is that paid units are taking more of the local results page every year, which raises the cost of not being there and the cost of being there badly.

What to do this week

Search the exact phrase a customer would type to find you. Note whether ads are showing at all. If nobody is advertising against a term, either there is no money in it or nobody has found it yet, and it is worth knowing which.

Then open your analytics and find how many people visited your site last month and how many contacted you. Run the five lines above with that ratio, your category cost per click, and your real close rate. It takes ten minutes and it usually ends the argument.

If the arithmetic does not clear, stop. Fix the page first. Ads will not rescue it, and running them now teaches you the wrong lesson about paid search.

Last, call your own business from a phone that is not in your contacts. Do it at 7pm on a weekday. What happens to that call is what happens to the calls you are about to buy.

Questions about whether ads make sense for you?

Email me at eric@seod.com with the three search terms you would want to buy, your average job value, and roughly what percentage of quotes you close. I will run the five lines and tell you whether the numbers clear, or whether the honest answer is not yet.

I answer these personally, and I will tell you to hold off if that is what the numbers say. Turning down work I do not think will pay back is cheaper for both of us than a client who quits in month three.

Or keep reading more on paid search for local businesses.