Tracking & Attribution

The £240,000 mistake that tracking would have prevented

In 2021, a fashion retailer spent £240,000 over 12 months on digital advertising. At the end of the year, the CEO asked a simple question: "Which campaigns actually generated revenue?"

The marketing team couldn't answer. They had impression data. They had click data. They had session data in Google Analytics. They had ROAS figures from each ad platform. But they hadn't installed conversion tracking on the website. They hadn't set up UTM parameters consistently. And different platforms were claiming credit for the same purchases through different attribution windows.

Best estimate: a third of their spend — roughly £80,000 — had gone to campaigns that generated little or no revenue.

Without tracking infrastructure, a marketing budget is a black box. Money goes in; some results come out; but the connection between the two is invisible.

(Illustrative scenario based on patterns common in retail digital advertising. Specific figures are representative of real-world outcomes — not a verified account of a specific named company.)

The tracking architecture

Effective paid advertising tracking has four components:

Each component serves a different purpose. Remove any one of them and you lose a dimension of visibility.

Platform pixels: the ad platform's tracking mechanism

Google Tag (Google Analytics + Ads): A snippet of code installed on every page of your website. Tracks:

• Which pages users visit
• How long they stay
• What actions they take (purchases, form submissions, phone calls)
• Where they came from (direct, organic search, paid search, etc.)

Google Tag Manager (GTM) is the recommended implementation method — a container that manages all tracking scripts without touching website code for each change.

Meta Pixel: Covered in the Meta Ads module. Tracks website behaviour and attributes it to Meta campaigns and ad sets. Also enables retargeting and lookalike audiences.

LinkedIn Insight Tag: For LinkedIn campaigns. Tracks website visitors and enables LinkedIn retargeting and conversion reporting.

The golden rule: Install tracking before running any campaign. Retroactive tracking installation means losing all data from campaigns already running.

UTM parameters: universal tracking

UTM parameters are tags added to URLs that tell Google Analytics (and most CRMs) exactly where a visitor came from.

The five UTM parameters:

A tagged URL looks like this: https://yourbrand.com/sale?utm_source=facebook&utm_medium=cpc&utm_campaign=spring-sale&utm_content=carousel-v2

When someone clicks this link, Google Analytics records: "a visitor came from Facebook, via a paid CPC campaign, in the Spring Sale campaign, via the carousel-v2 creative." Every conversion from that visitor is attributed to those parameters.

The UTM workflow:

1. Create a naming convention (consistent naming is critical — "facebook" vs "Facebook" vs "fb" are three different sources in GA4)
2. Use Google's UTM Campaign URL Builder (free tool) to generate tagged URLs
3. Apply tagged URLs to every ad, every link, every campaign
4. Never use untagged URLs in paid campaigns

UTM naming convention example:

utm_source: google / facebook / linkedin / email / tiktok
utm_medium: cpc / paid-social / email / organic
utm_campaign: [month]-[offer]: jan-free-consultation / spring-sale
utm_content: [format]-[version]: single-image-v1 / video-testimonial-v3

Conversion tracking: defining what success looks like

What is a conversion? A specific action on your website that represents a business outcome:

• Purchase (e-commerce)
• Lead form submission
• Phone call from the website
• Free trial sign-up
• PDF download (for lead generation)
• Booking or appointment

Setting up Google Ads conversion tracking:

1. In Google Ads: Tools → Measurement → Conversions → Create conversion action
2. Define the conversion (e.g., "Lead Form Submission")
3. Choose the method: tracking code on the "thank you" page, or Google Analytics import
4. Install the code (or import from GA4 — the simplest method)
5. Test with Google Tag Assistant to confirm it fires

Setting up Meta Pixel conversion events:

1. In Meta Events Manager: Create custom conversions or use standard events
2. Add the event code to the relevant page (e.g., the order confirmation page triggers a Purchase event)
3. Test with Meta's Pixel Helper Chrome extension
4. Use the conversion in your campaign as the optimisation event

Server-side tracking (advanced): Standard pixel-based tracking is increasingly limited by ad blockers, iOS privacy changes, and cookie restrictions. Server-side tracking (via Meta's Conversions API or Google's Enhanced Conversions) sends conversion data directly from your server — bypassing client-side limitations. For high-spend advertisers (£5,000+/month), server-side tracking can recover a meaningful proportion of conversions missed by pixel-based tracking — Meta and Google have cited recovery rates of 10–30% or more in their own documentation.

Attribution models: different ways to credit conversions

Last click: 100% credit goes to the final ad clicked before conversion. Simple; biased toward bottom-of-funnel channels (search, retargeting). Available in GA4 but no longer its default.

First click: 100% credit goes to the first ad the customer ever clicked. Biased toward top-of-funnel channels (social, display).

Linear: Equal credit distributed across all touchpoints. More holistic; requires complete journey data.

Time decay: More credit to touchpoints closer to the conversion. Reasonable for long consideration cycles.

Data-driven (GA4's default model): Machine learning assigns fractional credit based on which touchpoints statistically contributed to conversion. This is now the default in GA4 for properties with sufficient conversion volume. Requires minimum conversion volume to work reliably; properties with fewer conversions fall back to last-click.

The practical implication: GA4 defaults to data-driven attribution, which distributes credit across the customer journey rather than crediting only the final touchpoint. For campaign-level analysis, compare the data-driven view against last-click to understand where awareness channels (display, YouTube, top-of-funnel Meta) are being undervalued by simpler attribution.

Back to the fashion retailer

After 18 months and £240,000 in spend, the marketing team couldn't answer the CEO's question: which campaigns generated revenue? Fixing the tracking didn't change the campaigns — it changed every decision they made about the campaigns. Roughly £80,000 had likely gone to campaigns that generated little or no return, and they had no way to know which ones. Once conversion tracking, UTM parameters, and consistent attribution were in place, the data became actionable. The campaigns weren't the problem. The absence of measurement infrastructure meant they were optimising blind — and would have kept doing so indefinitely.

Key takeaways

• Tracking must be set up before any campaign launches. Retroactive installation means losing all data from campaigns already running.
• UTM parameters are the universal tracking layer. Tag every ad link consistently; use a documented naming convention; never send untagged paid traffic to your website.
• Conversion tracking tells the algorithm what to optimise for. Without defining conversions, smart bidding strategies optimise for clicks — not outcomes.
• Platform ROAS and GA4 will never perfectly agree. Different attribution models explain the gap. Use GA4 for neutral cross-channel comparison; use platform data for channel-specific optimisation.
• Server-side tracking recovers conversions that pixel tracking misses. For high-spend accounts, the investment in Conversions API (Meta) or Enhanced Conversions (Google) pays back quickly in better optimisation data.