Updating best-seller lists in near real-time


Most of us would assume that booksellers around the world were ready and waiting for news on this year’s winner of the Nobel Prize for literature, poised to promote his or her books as soon as the announcement was made. And whilst this may have been true, in reality it took a surprising amount of time for their websites to be updated, all, that is, for one – Adlibris.

Within five minutes of the announcement that Patrick Modiano had won, Adlibris’ best-seller lists were already reflecting the trend for customers searching for, and buying his books. In contrast, 30 minutes later other bookseller best-seller lists were yet to acknowledge the news.

Patrick Modiano

Stealing a march on the competition

For any such announcement there is a finite amount of demand generated by the news, and so it is imperative that retailers capture as much of that demand as possible before it wanes. In today’s competitive retail landscape every minute counts.

Clearly, for a best-seller list to first change, there has to be some initial external interest, i.e. interest generated from beyond the retailer’s shop walls. But the best merchandisers are those who recognise a trend early and use this initial demand to generate further interest from within their existing customer base. And an effective way of achieving this is through judicious use of best-seller lists. Nothing is more persuasive than peer recommendation.

Adlibris Best-seller-list a few minutes after the announcement that Patrick Modiano had won the Nobel Prize.

The problem with traditional best-seller lists
Most best-sellers lists are generated from sales during time periods that cover a day to up to a week. Such timeframes create accurate lists that reflect ‘the big picture’, but often there are other, smaller trends hidden in the sales data – trends that, unless specifically looked for, are destined to remain hidden in the lee of much larger waves. With most best-seller lists only exposing the ‘top ten’, it is inevitable that they get dominated by those trends whose life-span lasts longer than the timeframe in which the analysis is done.

The difficulty in uncovering smaller trends, those that happen over shorter periods or that manifest themselves within smaller cohorts of customers, is that it is much easier to get things wrong. By definition, there is less data to work with therefore it is harder to separate real signals from noise.

The solution
The key to solving the problem is knowing how to age data, such that sales relating to the bigger (obvious) trends can be down-weighted to allow the sales relating to smaller (less obvious) trends can come to the fore. The critical term here is ‘how’. What Apptus have discovered, and what has allowed the Adlibris best-seller list to work so successfully, is that the process by which sales data is aged depends on the context – there is no one ageing method that works universally. Knowing which method to use when is central to producing effective, reactive best-seller lists.

The benefits
Being the first to spot, and react to emerging trends is critical to successful retailing. But more than this, generating best-seller lists that react in near real-time creates customer experiences that feel fresh, informed and aware, and that will make customers value your brand and want to return.

To learn more about how Apptus can help with your merchandising challenges please visit apptus.com or get in touch.