“Is the website working?” usually gets answered with a feeling. It looks fine, nobody has complained, it’s been up for years. None of that tells you whether it brings in business. You can find out with two free Google tools and a short check once a month. Most of the work is setting things up properly once.
Step 1: write down the one action that counts
Every site exists to get someone to do something: send an enquiry, book a call, buy, phone you. Pick the main one. If there are two (a shop that also takes trade enquiries, say), pick both, but keep it short. Page views and followers are not on this list.
Then work out how that action shows up technically. A form submission, a visit to a thank-you page, a click on a tel: link, a completed checkout. If you can’t name how it would be recorded, you currently have no way of knowing whether the site works. The good news is that it’s fixable.
Step 2: make sure it is actually being counted
In Google Analytics 4 the actions that matter are called key events. Google defines a key event as “an event that measures an action that’s particularly important to the success of your business” (Google Analytics Help). Older guides call these conversions. GA4 now reserves “conversion” for actions shared with Google Ads.
Check three things:
- The event fires when the action happens. Submit your own form and watch it arrive in the Realtime report.
- It’s marked as a key event.
- It fires once per action, not on every page load of the thank-you page or every time someone hits back.
Also connect Google Search Console. Analytics tells you what people do on the site; Search Console tells you what they typed into Google before they arrived, and which of your pages Google shows for it.
If you’re on Shopify, sales are already recorded in the store’s own reports. You still want GA4 for where buyers came from.
Step 3: the monthly check
Once tracking works, these are the numbers to read, and the question each one answers.
- Key events for the month. Is the site producing the thing it exists for?
- Key events divided by sessions. Of the people who arrive, what share act?
- Traffic acquisition, by channel. Which sources send people who act. A big gap between sources is a marketing question more than a website one.
- Landing pages, with key events alongside. Which entry pages do their job and which ones people leave from.
- Engagement rate. A rough filter for visits that went nowhere.
- Search Console queries and clicks. What people searched to find you, and whether that matches what you sell.
A note on engagement rate, because GA4 treats bounces differently from the old Universal Analytics. GA4 counts a session as engaged if it lasts longer than 10 seconds, includes a key event, or has two or more page views. Bounce rate in GA4 is simply the share of sessions that weren’t engaged (Google Analytics Help). It’s useful for spotting a landing page nobody stays on. It’s not a goal in itself.
Reading it: a worked example
A fictional example to show the reasoning. A renovation company’s site had 1,400 sessions last month and 6 enquiries. That looks low, but the total hides the useful part. The landing pages report shows the bathroom renovation page had 500 sessions and 5 enquiries. The kitchen page had 450 sessions and none. Both get their visitors from a similar mix of search and ads.
So the site isn’t failing everywhere. One page is. The owner opens the kitchen page on a phone and finds the enquiry button scrolls off the bottom of the screen and the form asks for a floor plan upload. That’s the fix, and redesigning the whole site would have been money spent in the wrong place.
The numbers can point elsewhere, too. If one traffic source converts well and another hardly at all, the site is probably fine and the marketing is reaching the wrong people. Why traffic isn’t becoming customers covers that side. If every page and every source does badly, the problem is the site as a whole, and this walk-through of why good-looking sites don’t sell is the place to start.
Things that will mislead you
- One month on its own. Seasonal businesses swing. Compare with the same month last year when you have it, and look at the direction over three or four months.
- Small numbers. Going from 4 enquiries to 6 is not a 50% improvement you can bank on. With low volumes, wait for more data before deciding anything.
- Your own visits. If you and your staff use the site all day, you’re inflating the traffic and deflating the rate. GA4 lets you filter internal traffic by IP.
- Enquiries that never reach Analytics. Phone calls don’t show up unless you track taps on the number, and even then you miss people who dial it by hand. Ask new customers how they found you and write it down. It’s crude, and it catches what the tools miss.
What to do with the answer
If the numbers are healthy and steady, leave the site alone and spend your energy elsewhere. If they point at the site, fix the page they point at, then wait a month and read the same numbers again so you know whether the change worked.
If you’d like help setting up the tracking or reading what it shows, send us the site and what it’s meant to do. Our web service page explains how we plan a site around the action it needs to produce.