2017.04.25

How AdLibris is using AI to Power Sales Growth of up to 49%

Reading time: 6 minutes

Adlibris

Apptus eSales’ machine learning and continuous testing across search relevance and product visibility supports conversion rates that are three times higher than the market average.

About Adlibris:

  • Delivers 20 000 books every day to customers in Sweden, Norway and Finland
  • Founded in 1997 and is today the leading internet bookstore in Scandinavia
  • The business is head quartered in Stockholm, Sweden

Since its launch in 1997 Swedish online book retailer, Adlibris, has enjoyed astonishing growth. Today, it operates a catalogue that runs beyond 10 million products, serves three million visitors per month, and sells 300,000 products per day.

The challenge

With Apptus eSales already in place, Rubie and her team already had the ideal solution - an AI- powered, machine learning solution that could step in instantly and whose performance would only improve over time. Rubie explained:


“Apptus eSales AI-Powered eCommerce Optimisation provided the perfect solution for our visibility challenge. It immediately ensured that our product panels started recommending articles across categories. For example on a product page for knitting yarn the recommendations were for knitting needles, other yarn products and books about knitting. ”

“But as it learned more about user responses to different panels, so the recommendations improved. Now we have a fully edged automated cross selling recommendation system on the website - one that learns and tests as it goes and requires only minimal input.”

AI and the Future of Online Retail

As the eCommerce market continues to adapt to changing user behaviour and ever more demanding customers the challenges multiply, the argument for AI couldn’t be stronger. Rubie pointed out:

“Today, the big challenges are related to product discovery, relevance, search speed and personalising the user journey. And in response, AI is only going to become more important. These are not challenges that can be solved based on human input alone.” 

Adlibris offers a case in point. The business is growing fast and is constantly evolving its sales channels - and, according to Rubie, Apptus eSales plays an important role: “Apptus’ AI capability gives us the tools we need to seize the big opportunities. It enabled us to launch an eBookstore through our eReader, Letto, and provides the recommendation engine for personalised newsletters.

“Then there is our new app project. The speed of the Apptus eSales search engine increases app performance times signi cantly, which enhances the overall user experience.”

If there was ever any doubt, Adlibris clearly demonstrates to retailers that optimising their eCommerce operations with AI-powered merchandising pays - handsomely.

The solution

Revenue increased by 14% Conversion rate increased by 5%.

When asked what sparked the need to change their search, merchandising and personalisation solution, Rubie explains that their previous site had a basic search engine, which was configured using their own calculated sales and relevance statistics. Initially it worked well enough, but Adlibris' customers began to ask for better results that showcased more of thieir catalogue, not just the best selling titles. Furthermore, their company vision was evolving.

Instead of competing on just the usual aspects of price, customer service and fulfilment, we wanted to add a more personal experience and to inspire our customers in their choice of literature.

Adlibris looked at a few different companies but the most important criteria that made them select Apptus was because Apptus eSales was already widely used by other retailers, some of them in the same line of business, and, having been developed over a number of years, far more sophisticated than competing solutions.

We did not have the time nor resources to develop our in-house search engine to this level – we needed a product that worked out of the box.

The reason Adlibris disregarded other soultions was mainly the time and resources that would have been required to get them to the level already reached by Apptus.

At the start, our marketing department was a bit sceptical of using an algorithmic approach to selecting recommendations, preferring the manual approach we were used to. But we are steadily moving away from this method, allowing the system to make choices based on the customer, reserving manual selections for purely editorial material.


Adlibris have also gained more out of Apptus eSales by having regular contact between their Apptus account manager and their online merchandiser. This has had a major effect on their understanding of the solution and its fine tuning, how they plan feature requests and understanding each other’s businesses.

We have a tradition of building all our systems in-house, but we would never have had the resources and the focus to build a search engine as sophisticated as Apptus eSales. Working with such a dedicated team who’ve developed such a strong product has been very inspiring. This makes Apptus worth a look!

Apptus eSales has been invaluable in our journey, and central to achieving a conversion rate which is more than three times higher than the market average.

Rubie Oliphant

Product Specialist, AdLibris

Download the case as PDF

OPEN THE ADLIBRIS CASE