"Our analytics say one thing, our CRM says another." It's one of the most common problems we see — and it makes every marketing decision harder, because nobody trusts the numbers.
Google Analytics 4 (GA4) is powerful, but it's easy to configure in ways that quietly distort your data. These are the mistakes we encounter most often, and how to fix them.
A quick note on terminology: in GA4, the actions you mark as important are now called key events. The term "conversions" is used for events imported into Google Ads. The principles below apply to both.
1. Counting the same conversion twice
Duplicate tracking is extremely common. It happens when:
- The GA4 tag is installed twice — for example, hard-coded in the theme and added through Google Tag Manager.
- A form-submission event fires from both a tag and a plugin.
- A "thank you" page is counted every time it's reloaded or revisited.
How to check: use Tag Assistant or GA4's DebugView to submit a test form and watch how many events fire. For thank-you pages, consider firing the event on the actual successful submission instead.
2. Tracking clicks instead of outcomes
A click on a "Submit" button isn't a lead — the form might fail validation, or the user might abandon it. Tracking the successful outcome (a confirmed submission, a completed booking, a purchase) is far more reliable.
Where possible, trigger key events from a confirmation signal: a success message, a server response, or a data layer event pushed by the form itself.
3. Inconsistent event naming
When events are named ad hoc — form_submit, FormSubmit, contact-form, lead — reports become impossible to read and easy to misinterpret.
Agree on a short naming convention, document it, and use GA4's recommended event names (such as generate_lead and purchase) where they fit. A simple measurement plan listing every event, its trigger, and its parameters prevents most of this.
4. Not filtering internal and developer traffic
Your team, your agency, and your developers visit the site constantly — especially around launches. Without filters, that traffic inflates sessions and can trigger test conversions.
Define internal traffic rules in GA4's data stream settings, activate the internal traffic filter, and exclude test environments from your production property.
5. Broken cross-domain and payment-provider journeys
If your journey crosses domains — a separate booking system, a checkout on another domain, or a payment provider redirect — sessions can break apart. The conversion then gets attributed to the payment provider or appears as "direct" traffic.
Configure cross-domain measurement for domains you control, and add payment providers and similar services to the unwanted referrals list so they don't steal credit for conversions.
6. Misconfigured consent
Privacy regulations require consent for many types of tracking, particularly for visitors in the UK and European Economic Area. Two problems are common:
- Tags firing before consent, which creates compliance risk.
- Consent implemented so bluntly that measurement collapses without anyone realizing why numbers dropped.
Google's Consent Mode lets tags adjust their behaviour based on the visitor's choices. Since March 2024, Google has required advertisers using its advertising features with EEA traffic to send consent signals, including the ad_user_data and ad_personalization parameters introduced with Consent Mode v2. Work with your consent management platform to implement this properly, and have your legal adviser confirm your approach.
7. Never reconciling with the source of truth
Analytics will never match your CRM or e-commerce platform exactly — ad blockers, consent choices, and attribution models all create differences. But the gap should be understood and stable.
Once a month, compare key events in GA4 with the leads in your CRM or orders in your store. If the ratio suddenly changes, something has broken — often after a website update or a new plugin.
Build a measurement plan first
Most of these problems share a root cause: tracking was added piece by piece, without a plan. A simple measurement plan answers:
- Which business outcomes matter (leads, purchases, bookings)?
- Which events represent them, and what triggers each one?
- Which parameters should each event carry?
- Where does each event need to go — GA4, ad platforms, the CRM?
With that in place, implementation becomes straightforward and your data becomes something your team can actually rely on.
If your numbers don't add up, our digital growth & integrations service begins with an audit of exactly these issues.