Digital Marketing
Analytics & Tracking Setup
The least exciting month of any engagement and the one that makes the other eleven worth paying for.
Analytics and tracking setup makes your measurement trustworthy: conversions defined properly, events firing once, consent respected. The Nexclick does this before anything else, because every optimisation decision for the next year rests on whether these numbers are correct in the first place.
Is this you?
What usually prompts the call
- Your ad platform, analytics and CRM report three different conversion numbers.
- Form submissions are not tracked at all, or are counted twice.
- Nobody can say confidently which channel produced last month’s enquiries.
- Your tracking was set up years ago by someone who has since left.
What we do
The actual deliverables
Things that appear on an invoice, not adjectives.
- Audit what is currently firing
- Every tag, event and conversion, and whether it does what it claims. Duplicate events and conversions counted twice are the most common findings.
- Define conversions against business events
- A qualified enquiry, a booked call, a completed sale — not a pageview on a thank-you page. Definitions agreed in writing before implementation.
- Rebuild the tag container
- A clean Google Tag Manager setup with tags that fire once, on the right trigger, in the right order. Inherited containers accumulate tags nobody can account for.
- Consent mode implemented properly
- Nothing non-essential fires before opt-in, and consent signals reach the platforms correctly. A PECR requirement, and it materially affects what you can measure.
- Cross-platform reconciliation
- Why the ad platform, analytics and CRM disagree, and which to trust for which question. They will never match exactly, and the reasons are explainable.
- Offline conversion import where relevant
- Where the real sale happens on the phone or weeks later, fed back so the platforms optimise towards revenue rather than form fills.
- Documentation of the measurement model
- What each event means, when it fires and what it excludes. Without this, the next person inherits a container they cannot safely change.
Comparison
Why your platforms disagree, and which to believe
Three systems reporting three numbers is normal and explainable. Here is why each differs, and which to trust for which question — the goal is understanding, not forcing them to match.
| Discrepancy | Cause | Which to trust |
|---|---|---|
| Google Ads reports more conversions than GA4 | Ads uses click-through attribution over a longer window | Ads, for ad performance; GA4 for site behaviour |
| Meta reports far more than GA4 | Meta claims view-through conversions; GA4 does not | Neither alone — check against the CRM |
| GA4 sessions lower than server logs | Consent, ad blockers and JavaScript failures | Server logs for traffic volume |
| CRM has fewer leads than the form counter | Spam filtering, or failed CRM writes | CRM, and investigate the gap |
| CRM has more leads than analytics | Phone calls and direct enquiries | CRM — analytics never saw them |
| Conversions doubled overnight with no change | An event firing twice, usually after a deployment | Neither — fix it |
| Revenue differs between GA4 and the platform | Tax, shipping or refunds included differently | The platform, and align the definitions |
| Attribution disagrees on which channel gets credit | Different attribution models by design | Use one model consistently for decisions |
How it works
Step by step, with timeframes
Timeframes are typical rather than guaranteed, and they assume we get account access and approvals when we ask.
- 01Week 1
Audit
What fires, what it means, and where the numbers disagree. Produces a list of what is broken and what is merely misunderstood.
- 02Week 1–2
Define and agree
Conversion definitions written down and signed off, because most measurement disputes are definition disputes.
- 03Week 2–3
Implement
Container rebuilt, events configured, consent mode implemented, and everything tested in preview before publishing.
- 04Week 3–4
Verify and document
Live verification against real submissions, then written documentation of the measurement model.
What you get
Reporting and ownership
- A written definition of every conversion — what it means, when it fires, what it excludes.
- A clean tag container with nothing firing twice and nothing nobody can account for.
- Consent mode implemented so nothing non-essential runs before opt-in.
- A written explanation of why platforms disagree and which to trust for which question.
- Documentation sufficient for the next person to change it safely.
Tools and platforms
- Google Tag Manager
- Google Analytics 4
- Google Ads & Meta conversion APIs
- Consent management platform
- CRM integration for offline conversions
Timeline
How long this actually takes
Three to four weeks. It is the least interesting work we do and the highest-return, because everything after it depends on the numbers being right. Expect the audit to find something significant — a conversion counted twice, a form not tracked at all, or a definition that means different things in two systems. One honest note: platforms will never agree exactly. Attribution windows, cookie restrictions and modelling differ by design. The goal is understanding why they differ, not forcing them to match.
Pricing model
Fixed-price project
Fixed price, and the first work in almost every retainer. Can be bought alone if you want the foundations correct before deciding anything else.
Questions
Analytics & Tracking Setup questions
Why can our platforms not just agree?
Because they measure different things by design. Attribution windows, view-through counting, consent handling and modelling all differ between them. Chasing exact agreement is a waste of effort — understanding why each differs, and choosing which to use for which decision, is the achievable goal.
How much does consent affect what we can measure?
Meaningfully. Visitors who decline analytics cookies are not measured, and that proportion varies by audience. Consent mode allows platforms to model some of the gap. Anyone reporting figures without acknowledging this is presenting an incomplete picture as a complete one.
What is the most common problem you find?
A conversion counted twice — a tag firing on both a form submission event and the thank-you pageview. It inflates every downstream figure and makes cost per acquisition look better than it is, which then drives budget decisions in the wrong direction.
Do we need offline conversion tracking?
If the real sale happens on the phone or weeks after the form, yes. Otherwise the ad platforms optimise towards form fills rather than revenue, and they will happily find you more of the enquiries that never convert.
Could our own team set this up?
Basic event tracking, yes — the tooling is accessible. The parts that go wrong are deduplication, consent mode, cross-platform reconciliation and offline import. Those are where inherited setups are almost always broken, and where the errors are invisible until someone checks.
How is this different from GA4 implementation?
GA4 implementation configures one platform properly. This is the wider job — the tag container, every platform’s conversion tracking, consent, reconciliation and offline import. GA4 is a component of it, and it is the component people most often mean when they ask.
Last reviewed 28 July 2026.
Tell us what you are trying to fix
A 20-minute call, no pitch deck. The Nexclick will tell you what we would do, roughly what it costs, and whether we are the right people for it.