What recommender systems do today and in the future

by | 30 August 2018 | Basics

The variety of products that exist today, and which are presented to us in particular through online shopping, seems almost limitless. On the one hand, this is a great enrichment because it gives everyone the opportunity to buy exactly the product that best suits their needs. On the other hand, this diversity brings with it the challenge of having to orient oneself and compare products that are sometimes very similar. One possibility is to limit oneself to a certain brand or a certain solution. It is not without reason that we are increasingly seeing strong brands and the trend towards diversification and the emergence of closed ecosystems. But that alone is no longer enough today. The other solution that provides customers with orientation are Recommender systems. This data-driven approach provides enormous advantages and ensures an optimal customer journey.

Recommender systems as a prerequisite for next-best-action or next-offer marketing

The customer experience should be as optimal as possible. If this saves customers time and effort because they don't have to work their way through a myriad of different products themselves. A Recommender system is the basic prerequisite for Next-Best-Action-Marketing or next-offer marketing. This involves anticipating what the customer is most likely to be looking for and what is likely to be the next step in their customer journey. The goal is to reduce the bounce rate and maximise the conversion rate.

A recommender system can achieve this in different ways. The most important factor is the Data basis and the Forecasting methods that make the difference here.

Link tip: With these 5 measures to ensure optimal data quality.

Individual vs. collective recommender system

There are two types of recommender systems. On the one hand, so-called individual and on the other hand collective Recommender systems. On the one hand, the recommendation is based on the analysis of personal preferences of the individual customer. On the other hand, a recommendation can be based on the analysis of other customer decisions, i.e. a "collective" (Big Data). The underlying thesis is: customers who have bought a certain product are also interested in another product with a certain probability.

In principle, a third type of recommender system can also be mentioned here. For it is also possible to create a Hybrid form of individual and collective recommender System form. For both or all three types, it is important to note that they must be in line with the new General Data Protection Regulation. This means, for example, that data must be anonymised for or before analysis and customers must be asked for consent to use their data.

Improving the quality of decision-making

Recommender systems do not only exist in the area of product recommendation. Rather, they are only one of many examples of a recommendation system. In general, recommender systems can be defined as Toolsthat deal with large and complex amounts of information in order to be able to prioritise and facilitate decisions. Also Chatbots can take over this task, for example. Recommender systems can be used for any form of content. Three categories of decision bases can be distinguished:

  1. Information via the User (e.g. personal taste, purchase history etc.)
  2. Information about the Products, services and other objects themselves (e.g. food, films, books, technology, cars, etc.).
  3. Information via Transactions or similar Processes (e.g. recruiting, financial transactions, business decisions etc.)

The goal is always the same: the quality of the user's decision should be improved by the recommender systems. Data quality is essential for the functioning of recommender systems. If fake accounts or automated bots falsify the data basis, for example to push a certain product, the applicability of recommender systems is undermined.

Outlook: Wide range of applications for recommender systems

The diversity of recommender systems is correspondingly large. In the industries where these systems have been used so far, they have been largely responsible for the growth in these areas. Be it in:

  • Digital content for example, streaming services, newspapers, online magazines or social networks based on popularity,
  • Hotel recommendations on the basis of evaluations
  • or in the Retail and Trade for product recommendations based on shopping basket analyses

Future applications are conceivable wherever it is a question of the Improvement and Optimisation of the customer journey. Personalisation is also at the centre of the applicability of recommender systems, for example in the field of marketing. In the corporate context, knowledge databases can be sorted by relevance to facilitate access to important content. In particular, the combination of KI (artificial intelligence) and recommender systems are one of the most attractive innovations in this field. This makes them one of the most promising tools to effectively create value from data.

<a href="https://www.alexanderthamm.com/en/blog/author/michaela/" target="_self">Michaela Tiedemann</a>

Michaela Tiedemann

Michaela Tiedemann has been part of the Alexander Thamm GmbH team since the early start-up days. She has actively shaped the development from a fast-moving, spontaneous start-up to a successful company. With the founding of her own family, a whole new chapter began for Michaela Tiedemann at the same time. Hanging up her job, however, was out of the question for the new mother. Instead, she developed a strategy to reconcile her job as Chief Marketing Officer with her role as a mother.