How to Measure Whether Your Website Is Actually Working

Most small business owners have Google Analytics installed, and most of them have never logged in. The next most common group have logged in once, seen 47 numbers they don't recognise, closed the tab, and never gone back.

The question they actually want answered is simple: is my website working?

The good news is you don't need to understand 47 numbers. A handful is enough. This post walks through the ones that actually matter, in plain English, and the ones you can safely ignore.

First — define what "working" means for your business

We had a client tell us her website wasn't working, so we asked her what "working" meant. Turned out she wanted ten bookings a month through the site. Another client, same phrase, meant "getting my brand in front of people" — bookings weren't the point at all. A third one meant "ranking for my name on Google." All three were perfectly valid definitions of "working." But you can't measure a website against a goal you haven't written down. Before you open analytics, spend five minutes deciding what success actually looks like for your business.

For most small service businesses the answer is simple: enquiries per month. If your site is bringing you more enquiries than last month and they're becoming clients, it's working. If it isn't, it isn't.

Write your version of "working" down. Everything that follows is about measuring against that definition.

The four numbers that actually matter

1. Enquiries per month (the only number that pays the bills)

When we review a client's website performance, the first number we look at isn't traffic, rankings or bounce rate. It's enquiries. Everything else is context. A website that gets 100 visitors a month and produces 5 enquiries is working harder than one that gets 5,000 visitors and produces none. Traffic is a means to an end. The end is enquiries. If you can only track one number, track this one.

In GA4, this means setting up Key Events for:

• Contact form submissions
• Phone number clicks (on mobile especially)
• WhatsApp clicks
• Booking confirmations, if you've got an online booking system

If you haven't set these up yet, our guide to setting up GA4 walks through exactly how, and why it's the single most valuable 30 minutes you'll spend on your analytics.

What "good" looks like: Trending upward month on month. For most service businesses, a conversion rate of roughly 1–3% of relevant traffic into enquiries is a healthy baseline. If you're well below that, the site has a problem. If you're well above it, keep doing whatever's working.

2. Where your traffic is coming from

Traffic sources tell you where your visitors are coming from — Google, direct, social, paid ads, referrals. It's one of the most useful reports in GA4 and most people don't look at it.

What to look for:

• A healthy mix. If 95% of your traffic is from one source, you're dependent on it. If that source shifts — a Google algorithm update, a social platform's algorithm change, a Facebook ad account suspension — you're in trouble overnight.
• Quality, not just quantity. Traffic from Google usually converts far better than traffic from Facebook or Instagram. A drop in social traffic but a rise in Google traffic is usually good news even if the total number is similar.
• What's actually new. Is a source growing that you've done no work on? That often points to a page that's ranking for something unexpected, which is worth investigating.

3. Which pages are actually pulling their weight

This is the report that surprises clients most often. Go to the Landing Page report in GA4 — it shows you which pages people are entering your site on.

Most of the time, a client's homepage isn't the most-landed page. It's an old blog post, or a specific service page, or a location page they barely remember publishing. That's useful information. It tells you where your real entry points are — which means it tells you where to focus your time.

If a blog post from 2023 is bringing you 40% of your traffic, that post should probably have a clear CTA on it. If one service page drives most of your enquiries, the other service pages should probably be modelled on it. Let the data tell you where to spend effort.

4. Mobile vs desktop performance

Compare how your site performs on mobile vs desktop. In GA4, you can break most reports down by device category.

What to watch for:

• Mobile is usually 60%+ of traffic. If it's much lower than that, check that your site isn't actively pushing people away from mobile.
• Conversion rate on mobile vs desktop. If desktop converts at 3% and mobile at 0.5%, you've got a mobile experience problem — almost always slow load, hard-to-use forms, or CTAs that aren't where they need to be.

A mobile experience problem is often the single biggest fix available for a small business site. Our guide on how to get more enquiries without a redesign covers the specifics.

Numbers that look useful but usually aren't

Pageviews

The conversation we have most often with clients is about pageviews. "We had 12,000 pageviews last month!" Great — how many enquiries? "Oh, about the same as usual." That's the trap. Pageviews feel like success because the number is big and goes up. But pageviews don't pay the mortgage. We've had clients spend months chasing traffic growth who'd have been better off fixing the three things stopping their existing visitors from getting in touch. More traffic through a leaky funnel is just more leakage.

Bounce rate

Bounce rate is the metric we see small business owners misunderstand most often. The old definition — someone leaving without clicking elsewhere — made it look like a problem. But a visitor who lands on your contact page, reads it, then calls your number is a "bounce" in the old sense. GA4 changed how bounce rate is calculated entirely and the new number isn't directly comparable to the old one. If you're still obsessing over it as a headline metric, you're often measuring the wrong thing. Worry about whether they got in touch, not whether they clicked around first.

Session duration

Gets inflated by people leaving tabs open. A visitor who opens your site, gets a phone call, and forgets to close the tab for four hours shows up as a "four-hour session" in your data. Unless session duration is genuinely very short across the board — under 10 seconds average, which usually means the page doesn't load properly or doesn't answer the question — this number is mostly noise.

"Engagement rate"

Sounds meaningful. Rarely is. It's a composite metric that can move for a dozen reasons, most of which aren't useful to know. Focus on whether enquiries are going up.

How often should you actually check this?

Not daily. Daily analytics checks are counterproductive — the numbers are noisy day to day and you'll see patterns that don't exist. Most small businesses should look monthly. Ecommerce sites or anyone running active paid campaigns might want to look weekly.

Pick one day a month — the first Monday, the last Friday, whatever — and treat it as a standing appointment with your own data.

A simple monthly review that actually works

When we do a monthly review for a client, we don't open twenty tabs and dig into every report. We open four things: the last 30 days vs the 30 before, the traffic sources report, the landing page report, and the events/conversions report. Twenty minutes in those four views tells us 90% of what's actually going on. If we see something weird, we dig further. Most months we don't. A short, disciplined check beats an ambitious review you never actually do.

The rule we give clients: don't check analytics to feel informed — check them to make a decision. Every monthly review should end with you writing down one thing you're going to change, test, or investigate based on what you saw. Then come back the next month and see if it moved the needle. Most people treat analytics as something to look at. Treat it as a way to decide what to do next, and suddenly it earns its keep.

Here's the review, step by step:

1. Open GA4. Set the date range to the last 30 days vs the previous 30 days.
2. Check enquiries — up, down, or flat?
3. Check traffic sources — anything changed significantly?
4. Check landing pages — which got the most visits? Are they the pages you'd expect?
5. Check mobile vs desktop — is mobile converting at a similar rate to desktop?
6. Write down one thing you're going to do, based on what you saw.
7. Note the date. Come back next month.

That's it. Thirty minutes, once a month, done consistently, beats hours of dashboard-staring with no decisions attached.

When the numbers tell you something is wrong

Common patterns to watch for — and what they usually mean:

• Traffic up, enquiries flat. Conversion problem on-site. Copy, CTA, or form friction.
• Traffic falling across the board. SEO problem or an algorithm shift. Check if any of your main pages have dropped in rankings.
• Mobile traffic converting much worse than desktop. Mobile UX issue — speed, form usability, or CTAs below the fold.
• One page with an unusually high exit rate. That page isn't delivering on what visitors expected when they clicked through to it. Fix the headline, the first paragraph, or the CTA.
• Good traffic but the source suddenly shifts. Not always bad — sometimes a new page ranks. Worth investigating rather than ignoring.

You don't need to diagnose every problem yourself. But spotting the pattern is usually enough to know what kind of problem you have — and whether you need a developer, a copywriter, or someone to look at your SEO.

The bottom line

You don't need to become an analyst. You need a short list of numbers you check once a month, tied to a definition of "working" that makes sense for your business, with one action point at the end of each review.

For most small service businesses, that means: track enquiries, watch your traffic sources, keep an eye on which pages are pulling weight, and check your mobile experience is pulling its weight too. Ignore vanity metrics. Pick one thing to improve each month.

Do that consistently and you'll know exactly whether your website is working — and more importantly, what to do about it when it isn't.