Swain-Tech Studio websites for small businesses and independent creatives

Reading Your Own Analytics

Analytics dashboards are built for people running advertising campaigns. If you are a small business with a five-page site, ninety percent of what you are shown is irrelevant, and the effect is that most owners either stare at it anxiously or stop looking entirely.

Both are avoidable. Five numbers matter. Monthly is often enough. Here is the short version of what to look at and what to ignore. For remote teams that need explicit visibility rules, this article can help formalise what gets tracked.

The five numbers

How many people visited. The blunt one. Useful only as a trend across months, not as a figure — and a small site's month-to-month wobble is mostly noise.

How they arrived. Search, social, a link from somewhere, or typing your address. The most informative single number, because it tells you what is actually working. If most people arrive by search, your content is doing something; if most arrive from Instagram, the site is a landing pad for effort happening elsewhere.

Which pages they looked at. Frequently surprising. The page you agonised over may be barely visited, and a page you consider secondary may be the most read. That is information about what people want.

How many did the thing. Submitted the form, called, bought. The only number that connects the site to your income, and the one most people do not measure at all. For an independent reference beyond this site, Matomo is a useful place to compare approaches.

Phone or desktop. Decides where you spend effort. For most small businesses the answer is overwhelmingly phone, and it should change how you look at your own site.

What to ignore

Bounce rate. Widely misunderstood. Someone who lands on your contact page, reads your phone number and calls you counts as a bounce. That is a success.

Time on page. Measured badly and easy to misread — a long time may mean engagement or confusion.

Real-time visitors. Entertaining, useless.

Daily numbers. On a small site the noise exceeds the signal. Monthly, or quarterly.

Comparisons with industry averages. Your business is not the average of an industry, and the averages come from whoever is selling the report.

The question to ask instead

Not "how many visitors did I get" but "how many of the right ones, and did they do anything?"

Twenty visitors a month, three of whom enquire, is a better site than a thousand who read a blog post and leave. For a small business the numbers are small enough that this is legible if you look at the right thing.

Which is why the fourth number matters most, and why it needs setting up: count the form submissions, the calls, the orders. Without that, you are looking at traffic and guessing.

Setting it up so it stays useful

Record enquiries yourself as well. A simple note of where each one came from — asked on the phone, or a "how did you hear about us" field — beats any analytics package, because it captures what happened offline.

Mark the actions you care about in whichever tool you use, so the fourth number exists.

Look monthly, for ten minutes. Any more is anxiety.

Write down what changed. If you rebuilt a page in March, note it, so that a change in April can be attributed to something. The same log covers updates and renewals.

The privacy question, briefly

Analytics means a third party observing your visitors, and that carries obligations that vary by jurisdiction and by what you collect.

Two practical points. Consider whether you need it at all — if you would not act on the numbers, they are not worth the cookie banner. And consider a privacy-focused alternative: several lightweight tools give the five numbers above without personal tracking, no consent banner, and a fraction of the page weight. Third-party scripts have a real performance cost, and analytics is often the heaviest on a small site.

For anything health-related or otherwise sensitive, think harder about it than the default.

What analytics will never tell you

Why. It records what happened, not what someone wanted. A page with a high exit rate might be failing or might be answering the question completely.

Who did not arrive. The people who searched, saw your listing and chose someone else are invisible, and they are the interesting group.

Whether the enquiry was any good. Ten poor enquiries look better in a dashboard than two excellent ones.

So the most useful research tool remains asking people. How did you find me? What nearly stopped you getting in touch? Two questions on a first call, and they will tell you more about your site than a year of dashboards.

The short version