"How many products does an e-commerce shopper view in a typical session?"
A 6-month long study of 1.9M sessions of an e-commerce brand with shops across 5 EU countries, with an active catalog size of cca. 250; reveal that in almost a third of all sessions customers do not even view a product and in only around 3% of sessions do customers view more than 10 products.
This study has counted the number of view_items that occurred within each of the sessions and broke them all down by 5 intervals.
Table 1: In two thirds of all sessions, customers viewed one product or less.
These findings show us that these websites had an average conversion from starting a session to viewing a product at 67.23%, yet in approximately 88% sessions 3 or less products were viewed. Furthermore, we can notice that more than half, 52.65% of all viewed items occurred in only 11.35% of the sessions. This data shows that an e-shop visitor does not always view a product and if they do, they don’t view too many.
Therefore selection and merchandising of products is an important competency of an e-commerce firm, because if an e-shop has more than 200 active products per day; it is likely the shopper will only see a small part of the inventory at any given session. This is something we also understand intuitively, how many products do you view before making a purchase? How many times have you browsed more than 10 items in one session on an e-commerce store?
A common question that arises when discussing this analysis is that: the brand in the study had only 250 products, perhaps the customers simply didn’t have much choice to be clicking on? What if the shop had a far bigger selection on offer? Perhaps a larger selection would attract a customer to browse more?
To answer this question we conducted a similar, but shorter study of one month and 250k sessions. The e-commerce company in question had a product catalog 20,000+ strong and a price range between 1.50€ - 1000€, which yielded similar results. We can observe that the company was better on average at converting customers into viewing a product overall. (27.32% vs 32.77%)
Table 2: We discover a similar pattern in an e-commerce firm with a much richer catalog selection in terms of categories, price variation and sheer size.
From this data we can conclude most sessions do not result in many view_items and as such view_items are “rare”. Additionally, we can observe that the products customers do view also form their early opinion of your e-shop. In both examples of data, we can see that 1 product viewed per session is a big proportion of all sessions. With a large product selection, it is hard to know without a data-driven analysis, which product should have more or less of the exposure within the shop. If you would like to read about how to use a data-driven approach to systematically evaluate your products and decide which products should have more exposure and which ones less, you can do so here.