Google Ads for Small Business: A Practical First 90 Days, From Budget to Conversion Tracking

Start by writing down three numbers: what an average sale or job earns you after costs, what share of enquiries you close, and how many clicks it takes to produce an enquiry. Those three numbers decide what you can afford to pay for a click, and they matter more than any setting in the account. A small business that skips this step usually learns the numbers anyway, only with a month of ad spend to pay for the lesson.

The first 90 days then have three jobs: get measurement right, find the searches that produce real enquiries, and cut the ones that do not. It is a learning period with a budget cap, not a launch where results are owed on day one. If the business cannot answer the phone or reply to a form within a working day, fix that before paying for traffic.

Budget follows from your allowable cost per lead. Take the profit on an average sale, multiply by the share of leads you close, and that is the most a lead can cost before you lose money. Then set your target below that ceiling, because ads are not the only cost.

A worked illustration, with invented numbers. An electrician earns 500 in gross profit on the average job and closes about one lead in four. The break-even cost per lead is 500 times 0.25, which is 125. The electrician sets a working target of 60. If about one click in twelve becomes an enquiry, a lead at 60 implies a ceiling of roughly 5 per click. To collect ten to fifteen leads a month while learning, the monthly budget is about 600 to 900, so a daily budget near 20 to 30. If the market's click prices are well above 5, the business has learned something valuable before spending much: Google search may not fit at that margin, and a different channel or a higher-priced offer deserves a look.

Google's help page describes three match types. Exact match shows your ad on searches with the same meaning or intent as the keyword and gives the most control but the least reach. Phrase match covers searches that include the meaning of your keyword and reaches more than exact. Broad match reaches the widest range of related searches, including some that do not contain the direct meaning of your keyword, and Google says it should be used with Smart Bidding.

We recommend a cautious order for a new account. Begin with exact and phrase match on a short list of commercial-intent searches, such as the service plus your town, so that early spend goes to people who are plainly looking for what you sell. Once conversion tracking is returning dependable data, test a broad match keyword in its own campaign with a conversion-based bidding strategy and a modest budget, and judge it by the searches it actually matched. Broad match without conversion data to steer it is the setup most likely to waste a small budget.

Negative keywords stop your ad from showing on searches containing terms you list. Google's example is an optician adding 'wine glasses' and 'drinking glasses' because the word overlaps with the product. For a small business the obvious early negatives are job seekers, free, DIY, training, and the neighbouring services you do not offer.

One detail in Google's help page trips people up: negative keywords do not match close variants or expansions. Excluding the broad negative 'flowers' stops 'red flowers' but not 'red flower'. Singular and plural forms and synonyms need to be added yourself. Check the search terms report every week for the first two months and move the irrelevant queries into negatives; a short list kept current beats a long list written once.

Google's conversion measurement page explains that you choose what counts as valuable, such as a purchase, a sign-up or a phone call, and the platform then reports and optimises against it. For a service business that usually means three actions: a form submission, a call from the ad or a click on the number, and a call to the number on the website. Calls tracked through Google forwarding numbers can report duration and caller area code, which helps separate real enquiries from wrong numbers.

Counting every form fill as equal is a trap. If you have the capacity, record which leads became customers and import those results later, so the bidding strategy learns from sales and not from enquiries. Test the tracking before the first real click: submit your own form, place a test call, and confirm each shows up as a conversion. A campaign running on broken tracking is spending money to teach the algorithm the wrong lesson.

Think of three phases. In the first month the aim is a clean, small campaign with working tracking, and the right result is a handful of enquiries and a lot of search term data. In the second month you cut waste, rewrite ads against the searches that convert, and check that the lead handling is not losing people. In the third month you decide whether the arithmetic works and whether to scale, hold or stop.

Do not judge by clicks, impressions or Google's own optimisation score. Judge by cost per enquiry against the target you set and, where you can, cost per customer. If after ninety days cost per lead sits far above the break-even figure with clean tracking and sensible keywords, the useful conclusion is that this channel does not pay at your margin, and that is a decision you can make with evidence.

Google Ads is a poor fit when the numbers cannot support a paid click. If the average sale earns little, or clicks for your service are expensive relative to the profit, the arithmetic in the budget section will say so before you spend. It is also a poor fit for a business with nobody to answer the phone, for a product people rarely search for by name, and for a website that gives a visitor no reason to trust the business within a few seconds. In those cases, fix the business or choose another channel.

The usual failures are quiet. Accepting Google's suggested keywords and budget without checking them against your own numbers. Running ads to the home page. Leaving location settings on a wide area so that clicks come from people you cannot serve. Counting a click on a phone number the same as a connected call. Making many changes in the first week, so that no one can tell which change helped. A simple rule for a small account is to change one thing at a time and write it down with the date.

Finally, be honest about time. A small account needs about an hour a week for the first two months: search terms, negatives, tracking checks and lead outcomes. If nobody has that hour, hire help or hold the spend until someone does. An unattended account spends the budget the same either way, and it simply spends it with fewer corrections.