Marketing attribution is the process of assigning credit to the marketing touchpoints that contributed to a conversion. It sounds straightforward. In practice, it's one of the most contested and poorly understood areas in marketing — and getting it wrong costs businesses significant money in misallocated budget.
The problem with last-click attribution
Last-click attribution — the default in most analytics systems — assigns 100% of conversion credit to the final touchpoint before purchase. If someone found you through a blog post, followed you on LinkedIn, saw your Google retargeting ad, and then clicked a branded search ad before converting, last-click gives all credit to the branded search ad.
This systematically undercredits top-of-funnel channels (SEO, content, social, display) and overcredits bottom-of-funnel channels (branded search, retargeting). Businesses optimising on last-click data consistently under-invest in awareness and over-invest in capturing demand that their other channels created.
Attribution models: the spectrum
Single-touch models
- Last click: 100% credit to final touchpoint — default in most tools, systematically biased against top-funnel
- First click: 100% credit to first touchpoint — useful for understanding acquisition channels but ignores conversion path
Multi-touch models
- Linear: equal credit to all touchpoints — simple but treats an initial blog visit the same as a direct intent search
- Time decay: more credit to touchpoints closer to conversion — logical for short sales cycles, problematic for long ones
- Position-based (U-shaped): 40% to first touch, 40% to last touch, 20% distributed across the middle — values both acquisition and conversion
- Data-driven: uses machine learning to assign credit based on actual conversion patterns — most accurate but requires high conversion volumes
The measurement stack that actually works
No single attribution model is correct — each tells a different part of the story. A practical measurement stack:
- GA4 with data-driven attribution (requires sufficient conversion data) as your primary analytics model
- UTM parameters on all paid and email traffic — discipline here makes multi-touch analysis possible
- CRM-level attribution: ask new clients 'how did you hear about us?' and record the response — qualitative source data fills gaps that analytics miss
- Offline conversion import: connect phone call conversions and in-person sales back to online touchpoints
- Platform-level reporting: acknowledge it exists but treat it with scepticism — Meta and Google both take credit for conversions that would have happened anyway
The incrementality question: the most honest test
Attribution models tell you which channels touched a conversion. Incrementality testing tells you whether a channel caused it. An incrementality test holds back a random subset of your target audience from seeing a specific campaign, then compares conversion rates between exposed and unexposed groups. The difference is the true incremental effect.
This is the gold standard of marketing measurement, and it's increasingly accessible. Meta has a holdout testing tool. Google has Conversion Lift studies. Running periodic incrementality tests on your key channels is the most honest way to know whether your marketing budget is actually driving revenue.
Practical attribution for smaller budgets
For businesses without enterprise data science teams, a simple approach: (1) Use GA4 with data-driven attribution, (2) Ask every new customer how they found you and record it, (3) Review which channels appear in the path before high-value conversions using GA4's conversion paths report, (4) Run simple tests by pausing or scaling channels one at a time and observing business impact.
Attribution setup is part of every paid marketing engagement we run. We build tracking that gives you real insight into which channels are generating value — not just which ones are claiming credit.
Build my proposal