Predictive Pricing: Bookwire develops a self-learning tool for pricing eBooks and audiobooks

Frankfurt / Leipzig, 15 March 2018 – At Leipzig Book Fair 2018, the Frankfurt digital services provider announced a far-reaching technical innovation in the sphere of “big data” in the book industry: “Bookwire Predictive Pricing”. A self-learning recommendation mechanism has been developed, based on evaluation of publisher data. It automatically gives publishers recommendations for action on issues of pricing and marketing. The tool will be integrated into Bookwire MACS software in the summer.

“Bookwire Predictive Pricing” has been developed during the past six months by the Bookwire IT Team, together with a team of mathematicians and data scientists. The new tool is based on around 25 million datasets that are available to Bookwire from distribution and marketing for around 1,500 international publishing customers, over a period of eight years. They form the basis of the formula that is used for reliable and directly implementable sales prognoses for increasing turnover. 

Bookwire GmbH CEO Jens Klingelhöfer presented the new developed today (15 March 2018) to an audience of professionals at the “Zukunft.Verlag” symposium, at Leipzig Book Fair.

A new way of thinking: reliable forecasts in advance, instead of subsequent evaluation

“The clear objective was to turn around habitual behaviour in pricing or marketing campaigns for our publishers and, in so doing, open up entirely new opportunities. No more simply evaluating and drawing conclusions once a campaign has been executed, but predicting, based on the possibilities of artificial intelligence, which campaign and which price perfectly suit an individual title or title campaign, at which time. All with a very clear goal in mind: a direct and lasting increase in turnover”, according to Jens Klingelhöfer, CEO of Bookwire GmbH.

New technologies: research approach based on AI and machine learning

Along with the data material, the project team also used experiences from “Marketing Analytics”, which had already been successfully introduced in Bookwire MACS, and the expertise of a cooperation partner in the field of artificial intelligence, to develop “Predictive Pricing”. To begin with, nobody was sure how successful the research project would be.
“We weren’t sure whether it would be possible to calculate a formula that can be used accurately and in a way that publishers can immediately understand, from the content of the buzz words ‘big data’ and ‘artificial intelligence’ that are so popular, but often used inaccurately, in our industry”, continues Klingelhöfer. “The breakthrough came at the end of February 2018 and we will now integrate this new tool into Bookwire MACS by the summer.”

New benefits: optimal pricing lifecycles directly boost turnover

In concrete terms this means that “Bookwire Predictive Pricing” automatically actively suggests price changes or campaigns to publishers for every title in their catalogue. Based on the many data points collected by Bookwire, and other exogenous factors (e.g. weather or holidays), the formula calculates the ideal time for a price campaign, the most attractive price, recommends suitable marketing campaigns and shows which upsell to expect.

“This tool is ‘machine learning’ at the cutting edge of technology. The mechanism keeps learning and more successfully analysing the patterns of groups of buyers with every new transaction in Bookwire MACS – true artificial intelligence with knowledge that hardly any human beings would be able to provide. The issue of pricing is not a simple one in our industry, because fixed pricing for books means that the experience is simply not there. However, our publishers can now tackle the issue automatically. Everybody who works in marketing should be able to make full use of the potential of their own titles”, finishes Jens Klingelhöfer.