Discovery & Personalisation
Recommendations on PDP and basket
In one sentence
Other customers bought shows what shoppers with similar interest actually went on to purchase, appearing on the product page and basket to widen consideration beyond the single item in view.
How it works
This recommendation module surfaces products that other shoppers who viewed or bought the current item also purchased, shown on the product page and in the basket. Unlike a curated cross-sell, it's driven by aggregate purchase behaviour across the catalogue, so it surfaces genuinely popular alternatives and complements that a merchandiser might not have manually paired. It updates as buying patterns shift, and the brand controls where the module appears and can apply category or margin constraints to keep suggestions commercially sensible.
The problem it solves
For the brand
A shopper focused on one product is often unaware of other popular items the same audience buys — without a recommendation surfacing that, the retailer is relying entirely on the shopper's own browsing to discover them.
For their customers
Shoppers similar to them have effectively already done the research of what pairs well or what else is worth considering — without this signal being surfaced, that collective knowledge is invisible to a new visitor.
For shoppers
For the brand
Increases basket size through behaviourally-relevant recommendations
Average order value, recommendation click-through rate
Surfaces genuine demand patterns a merchandiser might miss
Cross-sell revenue attributable to recommendations
Recommendations adapt automatically as the catalogue and trends change
Recommendation freshness / relevance
Basket-stage suggestions capture late-funnel upsell opportunity
Basket-stage attach rate
In practice
A shopper views a running shoe product page.
The module shows insoles and running socks that other buyers of the same shoe also purchased.
At the basket stage, a further suggestion appears based on the overall basket contents.
The shopper adds an item they hadn't been looking for but recognise as genuinely useful.
Where it lands hardest
Health & nutrition
Effective for surfacing complementary formulations that shoppers with similar goals tend to combine.
Beauty & personal care
Strong fit — routine-based buying patterns produce genuinely useful behavioural recommendations here.
Food, drink & FMCG
Useful for surfacing flavour variety and pantry staples that co-occur in typical baskets.
Pet care
Good fit — owners of a given breed or life stage tend to buy recognisably similar product sets.
Fashion & apparel
Strong fit for outfit and styling-adjacent recommendations based on real co-purchase behaviour.
Luxury & premium
Honest note: lower basket volume gives the algorithm less signal to work with, so curated alternatives may outperform pure behavioural data.
Common questions & objections
Why the platform version wins
Because this reads live order and browsing data from the same platform rather than a bolted-on third-party recommendation widget, suggestions stay grounded in your actual shoppers' behaviour, not a generic model trained elsewhere.