Google Analytics 4 for small business: the reports that matter

Aug 22
3 min read

Updated: 3 days ago

Introduction

Google Analytics 4 is free, and for most small businesses it is also the least used tool they have installed. The interface changed, the old reports moved, and the sensible reaction was to stop opening it.

That is a shame, because four or five parts of it answer real questions. The rest can be ignored without loss.

1. Google Analytics 4 for small business starts with one decision

Decide what counts as a conversion before you look at anything else.

For most small businesses that is an enquiry, a booking, a purchase, or a call. Pick the ones that genuinely matter and mark them as key events. Everything downstream — which channel works, which page converts, whether spending is paying — depends on this being set correctly.

An analytics property with no conversion defined can only tell you about traffic, which is the least useful thing it knows.

2. Set up the events that matter, not all of them

GA4 collects a lot automatically: page views, scrolls, outbound clicks, file downloads. Useful context, none of it a business outcome.

What you add on top should be short: form submitted, call button tapped, booking completed, purchase. Mark those as key events so they appear in the conversion reports.

Resist tracking everything. Every extra event is noise in a report you are trying to read quickly, and the temptation is to admire the volume rather than act on the outcome.

3. The four reports worth opening

Traffic acquisition — where visitors came from, and which of those sources converted. This is the one that answers "is the channel working".

Landing page — which entry pages people arrive on and how they perform. This is where you find the page that gets traffic and converts nobody.

Conversions — your key events over time, with a comparison to the previous period.

Pages and screens — what is actually being read, which tells you what to write more of.

Four reports. If you open them monthly and act on them, GA4 has paid for itself several times over, and it cost nothing.

4. Compare periods, always

A number without a comparison supports no decision. Enquiries of 43 is neither good nor bad.

GA4 has a date comparison built into every report — use it on every look. Previous period for recent movement, same period last year for anything seasonal, which matters if your demand moves with terms, holidays or weather.

5. What to ignore

Bounce rate and engagement rate. Interesting, rarely decisive, and easy to misread on a single-page site.

Real-time. Compelling and useless. Nothing you should decide depends on who is on the site this minute.

Demographics and interests. Sparse for small volumes, and rarely changes what you would do.

Attribution model comparisons. Genuine value at scale, noise at a few hundred conversions a month.

6. Connect it to the rest of the picture

GA4 sees your website. It does not see your till, your phone, or your customer list — so on its own it cannot tell you what a customer is worth.

Export the pieces that matter into the same place as your sales data, so acquisition cost sits next to customer lifetime value. A free reporting tool will read both and refresh itself. That pairing is the decision; GA4 supplies one half of it.

7. Check that it is actually recording

The most common failure is silent. Tracking breaks during a site change, an event stops firing, and nobody notices for months.

Test your key events after any change to the site: submit the form, tap the button, and confirm the event appears. Add that check to whatever weekly routine you already run.

Analytics that quietly stopped working is worse than none, because decisions keep getting made on the last good month.

Conclusion

Define your conversions first, add a short list of key events, and open four reports rather than forty: traffic acquisition, landing pages, conversions and pages.

Always compare against a previous period, ignore the metrics that never change a decision, join GA4 to your sales data so cost and value sit side by side, and test that events still fire after any site change.