How Personalisation with Nosto Increased AOV by 28%

A multi-brand beauty retailer carrying 800+ SKUs was treating every visitor identically: the same homepage hero, the same featured products, the same 'you might also like' recommendations regardless of browsing history. With 50,000 monthly visitors and a broad product catalogue spanning skincare, haircare, and colour cosmetics, the missed personalisation opportunity was significant. They'd heard about Nosto but weren't sure if the investment was justified.

The Business Case

Before recommending Nosto, we ran three weeks of baseline measurement. We tracked which product category each session was primarily engaging with and cross-referenced with what was actually surfaced to them on the homepage and in recommendations. The data showed that 62% of visitors were seeing featured products from categories they had never interacted with — a direct relevance problem.

Nosto Implementation

We integrated Nosto with the Shopify store over four weeks. The integration used Nosto's Shopify app, supplemented by custom event tracking for category affinity scoring. We placed personalisation slots across six key locations in the customer journey.

• Homepage hero product carousel: personalised by category affinity
• Collection page: personalised product ranking within results
• Product page: 'complete the routine' recommendations based on product category
• Cart drawer: cross-sell recommendations based on cart contents
• Post-purchase email: product recommendations in day-7 follow-up
• Exit-intent popup: personalised offer based on browsed category

Category Affinity Logic

Nosto's default recommendation engine works well, but we customised the affinity model to account for the brand's specific category structure. A customer browsing retinol serums would be classified as 'skincare, anti-ageing' — triggering a homepage experience centred on that category's hero products rather than the generic bestseller carousel.

Measuring Incrementality

We ran Nosto's built-in A/B testing with a 20% holdout group seeing non-personalised recommendations for the first 60 days. This gave us a clean incrementality measurement rather than relying on click-attributed revenue, which can overstate personalisation's contribution.

Results

• AOV: £47 → £60 (+28%)
• Incremental annual revenue: £195,000
• Cart cross-sell click rate: 18% (vs 4% on previous static recommendations)
• Exit-intent personalised offer conversion rate: 6.2%