A beginner's guide to setting up web analytics for a UK small business

You have a website. You probably have some traffic. What you do not have is a clear idea of whether any of it is doing anything useful. That is the gap analytics is meant to close, and it is usually where small businesses get stuck: you sign into a tool, see forty numbers you never asked for, and quietly close the tab.

Setting up analytics for a UK small business does not need a data team, a dashboard designer, or a six-week project. It needs three decisions, one tag, and a routine you can keep to twenty minutes a month.

Start with the question, not the tool

Before you open any analytics product, write one sentence describing what the website is for. Not "to raise awareness" — something checkable. A Bristol heating engineer might write: the site should generate quote requests from homeowners within twenty miles. An online shop selling refillable cleaning products might write: the site should take orders without me answering a single email.

Next, list the actions that count as success. A phone call, a form submission, an online order, a booking, a newsletter sign-up, a brochure download. Once that list exists, you know both what to measure and what to ignore. Page views are an ingredient, not a result.

Choose a handful of metrics that earn their place

There are two common traps. The first is vanity metrics: page views and social likes that rise and fall without telling you whether anyone bought anything. The second is measuring everything, which produces a dashboard nobody opens twice. Aim for four families of numbers instead.

• Acquisition — where visitors came from: organic search, paid ads, email, social, referrals, or typed straight in.
• Engagement — which pages hold attention and which send people away, especially your service and pricing pages.
• Conversion — the actions from your list, counted by type rather than lumped together.
• Retention — returning visitors, repeat orders, and whether your email list is growing.

A starter set that fits on one screen

Sessions, conversions by type, conversion rate, your five busiest landing pages, channel breakdown, and the split between mobile and desktop. Six numbers, and you can explain every one of them to someone else. Add more only when a real question demands it.

Install one tag, and install it once

Google Analytics 4 is the common free default, and it is fine to start there. If you would rather avoid Google, tools such as Plausible, Matomo and Fathom offer lighter, privacy-focused alternatives, usually for a monthly fee. Whichever you pick, the installation process is broadly the same.

1. Create an account and a property, then add a data stream for your domain. Set the time zone to London and the currency to pounds, otherwise your reports will look wrong forever.
2. Copy the measurement ID for the stream — it looks like G-XXXXXXXXXX.
3. Install it through your platform's own route where possible: a WordPress plugin, a Shopify app, or the tracking field in Squarespace or Wix. If you need more control, use a tag manager and publish a container.
4. Open the realtime report and visit your own site in a private browser window on your phone.
5. Confirm you can see that visit. If not, check your caching plugin, your cookie banner, and whether an ad blocker is running.

One warning worth heeding: do not paste the same tag into the theme header and a plugin. Duplicate tags double-count your traffic, and the resulting mess takes an afternoon to untangle. Keep a short note of what you installed, where, and on what date.

Mark the moments that matter

A tag tells you someone arrived. Events tell you what they did. Set up an event for each action on your success list: form submissions, clicks on a phone number, checkout steps, booking confirmations, downloads. Most platforms have a built-in way to flag these as conversions, so they appear in reports rather than being buried in a debug view.

Name events so your future self understands them. quote_form_submit is useful in eighteen months; event_17 is not. Test every one of them by completing the action yourself and watching the realtime report. An event you have not tested is a guess.

Cut the noise before you trust the numbers

Raw data lies a little. Fix the easy problems first.

• Exclude your own visits. Your office IP address or a filter for logged-in staff traffic stops you inflating your own stats.
• Watch for referral spam — odd domains sending traffic that never behaves like humans.
• Check that your cookie banner actually blocks non-essential tracking until consent is given. Under UK rules, analytics cookies generally need consent before they are set, and most tools have a consent mode to handle this. If you are unsure how your setup stands, ask a qualified adviser.

Accept that analytics always undercounts. Ad blockers, privacy browsers and declined consent mean the real figure is a little higher than the report. Trends matter far more than exact totals, so do not chase the missing few per cent.

Build a dashboard you will actually open

One page, no scrolling, compared against the previous period. A free tool such as Looker Studio can pull straight from your analytics account, but the built-in reports or a monthly spreadsheet export are perfectly respectable choices. Arrange it like this:

• Top row: sessions, conversions, conversion rate.
• Second row: channels, so you can see paid versus organic versus email.
• Third row: top landing pages, plus the mobile and desktop split.

Resist adding a chart for every question you can imagine. If a number never changes a decision, it does not belong on the page.

Twenty minutes a month is enough

Pick a fixed date — the first working day of the month works well — and give it twenty minutes. Scan the dashboard, then write three lines: what changed, what you think caused it, and one thing you will try next month. One action, not five. A slow landing page on mobile, a form that asks for too much, a channel quietly growing: any of these is a good use of an hour's work.

Review the whole setup once a quarter and change a single metric if your business has moved on. Analytics rewards consistency far more than sophistication. A modest routine you keep beats an elaborate system you abandon in March. Numbers you never act on are just decoration.

Photo: ZigmarsBerzins / Pixabay