How AI can help list rankings to increase sales

Optimum ranking sequences 

Determining the best ranking sequence on an e-commerce site can deliver amazing improvements on sales. It requires either a crystal ball or some very smart mathematics! The fact is, given the variables and numbers of products and sessions, the challenge is best suited to be done by technology, as only technology has the capacity to continuously analyse behavioural data to keep the lists right up to the minute.

Also, because this smart technology is capable of recognising trends, i.e. when products are becoming popular, and it works 24 x7 so, even late at night, if a TV advert drives traffic to your site, the trend will be addressed, and the ranking sequences adjusted accordingly in real-time.

The traditional way of ranking lists is often based on all the sales numbers over a fixed time interval, say 3 months. This method works well only if your goal is to display the top sellers from the past 3 months, for example. 

But as soon as you want to restrict your list to a narrow category, this simple technique will miss products bought outside the timeframe, as it does not account for trends that may be on the decline.

The Apptus ranking algorithm handles the complexity

Products sell at different rates, which contributes to the difficulty in calculating the sequence, especially when this is being done in near real time.

Apptus_Webinar_16_9_Purple

This is a statistically complex problem, and we adopt what mathematicians refer to as the “half-life” concept to assess the sales distribution of all products.

The half-life method is well established in computer science; however, its implementation is far from trivial.

The good news is that we have used this approach to create an algorithm that can determine the best ranking sequence for any list of products, be they a search results list, a category list, or a filtered list. The sequence will be based on an overall business objective and consider trending and hot selling products whilst maintaining a good position for consistent sellers.

 

Want to learn more? Contact our team.

 

Email us