We inherited an account last year where the previous setup counted every single page view, including the cookie banner acceptance, as a "conversion." Six months of reporting had told the client their site was converting beautifully. It wasn't converting at all. Nobody had actually checked what the goal was measuring, because once a tracking setup looks like it's working, almost nobody goes back and questions it.
Why GA4 Trips People Up More Than Universal Analytics Ever Did
We get asked, fairly often, whether the move to GA4 was actually an improvement at all, given how many setups we've seen go wrong since the transition. Our honest answer is yes, eventually, once set up properly, because the event-based model genuinely captures user behaviour in more useful detail than the old pageview-centric system ever could. The transition period has simply been messier than most businesses expected, and plenty of migrations were rushed through without anyone checking the new numbers against the old ones for a sanity check.
Universal Analytics, whatever its flaws, made goals relatively hard to get catastrophically wrong — you set a destination URL or an event and that was broadly it. GA4 works on events as the base unit for everything, which is more flexible and genuinely more powerful, but it also means it's entirely possible to mark something as a "key event" (GA4's current term for a conversion) without fully understanding what that event is actually firing on. We see this constantly: a form submission event that fires on page load instead of successful submission, or a "contact" conversion that counts anyone who merely scrolls past a phone number.
The flexibility is the whole appeal of GA4. It's also exactly why so many small business setups in Essex are quietly measuring nothing useful, while the dashboard looks entirely fine. We've opened accounts where every number looked healthy and growing, right up until we traced a single key event back to its trigger and found it fired on a page that no genuine customer would ever have reason to land on.
Start With the Business Question, Not the Tool
Before opening GA4 at all, we sit down with a client and ask what actually counts as a result. For a driving school in Billericay, that's a lesson enquiry form or a phone click. For a solicitor in Thurrock, it's a callback request or a specific case-type enquiry, not a generic "contact us" click from someone reading a blog post out of curiosity. For an ecommerce client, it's obviously a completed purchase, but it's also worth tracking add-to-basket separately, because the gap between the two tells you something about checkout friction that purchases alone never will.
Skipping this step is the single biggest reason setups go wrong. Agencies and in-house teams alike tend to open GA4, see a list of suggested events, tick a few that sound relevant, and move on. That produces data. It rarely produces data that answers the question the business actually needs answered.
The Technical Bit, Kept Mercifully Short
Once the real goals are defined, the actual implementation is fairly mechanical:
- Set up enhanced measurement correctly, then turn off the bits that create noise (scroll tracking on a one-page site is rarely useful, for instance)
- Build custom events for the things that matter specifically to this business — not the generic defaults
- Mark only the genuinely meaningful events as key events, not everything that technically fires
- Test every single one manually, in an incognito window, before trusting a single number it produces
That last point gets skipped more often than it should. Testing takes twenty minutes. Discovering six months later that a form confirmation never actually fired, because the thank-you page URL changed during a site redesign nobody flagged, costs a lot more than twenty minutes of confidence in the meantime.
A Real Example of Getting It Wrong the First Time
We set up conversion tracking for a landscaping business near Epping a few years back and, in our defence, it looked correct. Form submission event, destination confirmation page, everything configured by the book. What we hadn't accounted for was that their contact form used JavaScript validation that occasionally let people submit twice if the page lagged slightly on a slow mobile connection, which happens a fair bit around some of the more rural parts of that area. We were overcounting conversions by roughly fifteen percent for two months before a sharp-eyed look at session-to-conversion ratios flagged it as implausible. We'd set it up "properly" and it was still wrong, because testing in ideal conditions on a fast office connection isn't the same as testing like an actual customer on patchy signal in a van.
That episode changed how we test now. We deliberately throttle our own connection during testing to mimic a poor mobile signal, specifically because that single landscaping client taught us an office test is nowhere near enough.
We've since added a rule to our own onboarding checklist: no account goes live for reporting purposes until someone on the team has submitted every tracked form from a genuinely slow connection, not just a fast office one. It's a small, slightly tedious step. It has caught at least two more broken events since we introduced it, both of which would otherwise have sat unnoticed for months, quietly feeding false confidence into a client's monthly numbers.
Why Doing This Early Saves Real Money
We'd rather spend an extra week confirming goal tracking is accurate before any strategy work begins than build months of recommendations on numbers nobody has verified. Every month that passes with bad goal tracking is a month of decisions made on fiction. Budget gets allocated toward whichever channel appears to be converting, content gets prioritised based on which pages seem to drive results, and none of it means anything if the underlying event is broken. We've seen clients pour spend into paid search for months because GA4 showed it outperforming organic, when the actual issue was that the organic-driven form simply wasn't firing its event correctly. Fixing tracking after the fact doesn't just cost setup time — it costs however many months of wrong decisions happened in between.
What Proper Setup Actually Looks Like on Day One
For a new client, we won't touch content strategy or technical SEO recommendations until goal tracking is confirmed working, because every single other decision downstream depends on it being accurate. That order of operations frustrates some clients who want to see immediate activity. We'd rather lose a week at the start than build a quarter of strategy on numbers nobody tested. This groundwork sits underneath every bit of ongoing analytics and SEO work we do after it, and skipping it tends to show up as confusion three or four months down the line, right when it's hardest to untangle. A client who trusts their own dashboard is a client who makes faster, better decisions with us, and that trust has to start on day one, not after a problem is already hiding in the data.
We've turned down starting content work early for a new client before, even when they were keen to get going immediately, purely because tracking wasn't confirmed yet. It's an awkward conversation to have in week one. It's a far easier one than explaining in month four why the results we reported never quite matched what they were seeing in their own bank account.
The Bottom Line
GA4's flexibility is a genuine strength, but it means a goal setup can look complete and still measure the wrong thing entirely, sometimes for months, without any obvious warning sign. Defining what actually counts as a result before touching the tool, testing every event manually in realistic conditions, and resisting the urge to tick the easy default options will save far more time than it costs. Get it right once at the start and every report after that is worth trusting. Get it wrong and you're making decisions on numbers that only look like evidence.
Looking for SEO agency in Essex?
Let's talk about what your business actually needs
Digital Impact Ventures provides SEO, AI visibility and digital marketing support for businesses across Colchester and Essex. Get in touch and we will talk through where your business is now and where it needs to go.