What to actually measure on a small business website

The right starting question is not “what can GA4 measure” but “what decision would change if I knew this number” — and for most small Australian businesses, that narrows a dashboard with dozens of available metrics down to a handful that actually matter.

Start from the business objective, not the dashboard

GA4 can report on scroll depth, video engagement, file downloads, outbound clicks and a long list of other automatically captured events, and it can be configured with named “goals” (conversion events, in GA4’s current terminology) for almost any of them. Most of those possible goals answer no question a small business owner actually has. The useful exercise is working backwards from a real business objective: name the two or three decisions this data could inform — should I keep spending on this ad campaign, is the new page working, are enough of the right people finding the site — and then find the specific number that answers each one, rather than starting from the list of everything the tool happens to track.

The small set that covers most businesses

What to track	The decision it informs	Where it lives
Enquiries (form submissions and, where relevant, phone calls)	Is the site producing the actual outcome the business needs	GA4 conversion events, see conversion tracking and goals in GA4
Which channel those enquiries come from (organic search, paid, social media, direct, referral)	Where to keep spending time or money, and where to stop	GA4 Acquisition reports, see traffic sources explained
Which queries bring organic visitors	Whether the content matches what people actually search	Search Console Performance report, see what Google Search Console shows that GA4 doesn’t
Device split (mobile vs desktop)	Whether a mobile problem is masquerading as something else	GA4 Tech reports
Trend over time on the above, not a single snapshot	Whether something changed, and when	Any GA4 report, compared across date ranges

What this deliberately leaves out, and why

Page views alone, session count alone, bounce rate alone, and time-on-site alone are not on the list above, not because they are hidden or hard to find, but because none of them, on its own, indicates whether the business got anything out of the visit — a low bounce rate on a page nobody actually converts from is not, by itself, a business result. They are covered specifically, and why they mislead when treated as a headline number, on vanity metrics and what to look at instead.

Different businesses genuinely need different things

A phone-first emergency trade cares more about tracked calls than form submissions, because the form is the secondary channel for that business — see call tracking for phone-first businesses. An ecommerce store cares about revenue, average order value and cart abandonment, none of which apply to a business that does not sell online. A professional services firm often cares more about enquiry quality — did the person who submitted the form actually match the service offered — than about raw enquiry volume, which is a harder thing to measure automatically and usually requires the business itself to tag and review incoming enquiries rather than relying on GA4 alone.

A worked example of applying the decision-first test

Consider a business deciding whether to keep running a paid advertising campaign next quarter. The relevant decision is specifically “does this campaign produce enquiries at a cost the business can sustain”, which points directly to two numbers: conversions attributed to that specific campaign, and how that figure compares against the campaign’s cost over the same period — not session count for the campaign, not the campaign’s click-through rate in isolation, and not a general site-wide traffic trend that would look identical whether or not the specific campaign was working. Naming the actual decision first, before opening any report, is what narrows a dashboard full of available numbers down to the two that genuinely answer the question being asked.

Under-measuring is a real risk too, not just over-measuring

While this page argues against tracking everything GA4 happens to offer, the opposite failure is equally real and arguably more common among small businesses: looking only at total visits or a single top-line traffic number, with no conversion tracking configured at all, and having genuinely no way to answer any of the decisions in the table above. A business in this position is not avoiding vanity metrics through restraint — it has simply not yet done the setup work covered in setting up GA4: the first week checklist that would let it measure anything more useful, and the fix is completing that setup, not avoiding measurement altogether.

A practical way to choose

Write down the two or three business decisions that data might change over the next quarter — a marketing spend decision, a “should I redesign this page” decision, a “is this new service page working” decision — before opening GA4 at all. Then find the specific metric under each of those decisions, ignore everything else in the interface for now, and revisit the list again next quarter as the decisions change. This produces a small, stable set of numbers worth checking regularly, instead of a large dashboard nobody reliably reads.

What to do next

If GA4 is not yet configured to track the events in the table above, GA4 basics for a non-analyst covers the tool itself, and setting up GA4: the first week checklist is the practical configuration list. For how this connects to what a small business site actually needs to do, what a small business actually needs sets the broader context this measurement serves.