Many companies involved in e-commerce spend too many man-hours per month manually merchandising their online store. They spend time associating products under headings such as “you may like” or “goes well with”. The problem with this is, that as well as costing time and money, any recommended products are static and are based only on what the merchandiser “thinks” will be relevant.
Other eCommerce sites use collaborative filtering, “the power of the crowd”, to relate products, for example “most popular” and basic “people who bought this also bought…” suggestions. This type of basic recommendation has its place on product pages – where visitors have shown an interest in a certain products with certain attributes, and it is logical to assume that the suggestions will be more relevant than not.
However, on landing pages and throughout the shopping experience, visitors aren’t always sure what they’re looking for.
Giving basic suggestions based on what other people have bought may interrupt their search, even turn them away in those crucial few seconds when they arrive on the page. By personalising the content and products shown as a result of tracking the behaviour of the visitor more relevant suggestions can be made. Behaviour such as the referring site or the referring search query can give instant insight into what the visitor is looking for; subsequent clicks a user takes around a site can present a deeper understanding of what the user is interested in.
For example, if a visitor arrives from Google after searching for “maxi dresses”, we can instantly show the visitor a selection of products based on what others who arrived the same way have done. This is different from collaborative filtering as it takes into account both visitor profiling and search analysis, rather than a broad brush segmentation approach.
If the same visitor looks at several dresses of the same colours, we can determine that the visitor likes products in those colours, and can adapt the suggestions for both the current and future visits.
It’s the small, but very clever, adaptations to suggestions which can increase conversions, increase average order values and increase returning visitor purchases.
No related posts.