Put this into practice
Run A/B tests and personalize your store without code. 14-day free trial, no credit card.
Start free trial →Run A/B tests and personalize your store without code. 14-day free trial, no credit card.
Start free trial →Last click attribution is an attribution model that gives 100% of the credit for a conversion to the final marketing touchpoint a customer interacted with immediately before completing a purchase or other conversion event. If a customer first discovered a brand through a Meta ad, later searched for the brand on Google, and then clicked a Google Ads branded keyword before buying, last click attribution assigns the entire sale to the Google Ads branded search — ignoring the Meta ad and organic search entirely.
Last click attribution is the default model in most ad platforms and Google Analytics (prior to GA4's shift to data-driven). Because it's the default, it shapes how most D2C brands allocate their marketing budgets — often incorrectly.
The appeal of last-click is its simplicity: one channel gets credit, reporting is clean, and there's no debate about shared credit. The problem is that it systematically undervalues awareness and mid-funnel channels (Meta discovery ads, YouTube, influencer content, email newsletters) and over-credits retargeting and branded search, which often just capture demand that other channels created.
For a brand spending ₹50,000/month on Instagram awareness ads and ₹30,000/month on Google retargeting, last-click will show the retargeting producing 8x ROAS and Instagram producing 1.2x. A naive reading of that data cuts Instagram — which then starves the retargeting pool of new prospects, and ROAS falls across the board. The root cause was the attribution model, not channel performance.
Consider a D2C supplements brand in India running campaigns across YouTube, Instagram, and Google Shopping. A customer watches a YouTube pre-roll for a protein powder, sees an Instagram feed ad two days later, searches "whey protein 1kg price" on Google, clicks a Shopping ad, and buys. Under last click attribution: Google Shopping gets 100% of the ₹1,800 order. YouTube and Instagram get zero credit. The brand cuts YouTube spend. New customer acquisition slows. The Google Shopping ROAS then drops because there are fewer new prospects entering the funnel. The attribution model created a decision that damaged the business.
Last click attribution can distort the apparent impact of landing page tests if different traffic sources have different closing propensities. A variant that happens to receive more last-touch traffic from high-converting branded search may appear to win even if the page itself is no different. Always segment test results by traffic source when diagnosing unexpected outcomes.
Run smarter A/B tests with CustomFit.ai — 14-day free trial, no credit card required.