salesFor marketing projects, customers or prospects are typically segmented into groups, such as by gender or geographic region, and automated routes are built from those segments. A really personalized approach, however, isn’t possible that way, only recommendations, offers or purchase proposals based on a cluster. Well, not only does this practice have high waste circulation, it regularly upsets customers when, for example, they are offered things they have long since bought.

Although the necessary information for a truly personalized, individual customer approach is usually available, the lack of resources and tools for harnessing this data remain challenging: the database is simply too complex to make a manual setup of personalized automation routes possible.

Here, the use of artificial intelligence (AI) creates completely new possibilities. Rather than manually segmenting by specific groups of buyers, creating the content with the appropriate fixed incentives, and then monitoring and, if necessary, controlling the effectiveness of the campaign, the marketer can focus on basic strategic decisions - with the remainder being provided by artificial intelligence.

Based on the strategic specifications, AI decides based on patterns pulled from customer data individually about the time, the channel, the content and the type of incentives.

For instance, an AI-based incentive recommendation engine can calculate for each individual customer how likely that person will be to place an order shortly in the next few hours or days. On this basis, the algorithm decides whether an incentive is necessary and, if so, to what extent. If it is a customer who buys very regularly, one can assume that he/she will make an order very soon, hence, an incentive is not necessary. Likewise, the AI can help reactivate lost customers by providing longer inactive contacts with a higher incentive, accepting a lower margin in favor of the reactivation.

Another example: if buyers cannot be reached via email, it may be possible through Facebook or Google Ads. Since Facebook and Google have an average matching rate of 50 to 70 percent, these additional channels offer a much wider range. In this case, an adjustment takes place based on the email address. If the recipient is registered at Google or Facebook with the same email, matching will take place there. In order to comply with the privacy policy, the matching happens based on pseudonymous data. A transfer of personal data to third parties, in this case Facebook or Google, does not take place.

With the help of AI, Facebook and Google cannot only reactivate inactive customers, but also acquire new customers. For this purpose, ads are automatically imported into Facebook or Google based on the behavioral and lifecycle data of users. In other words, it is possible to select VIP customers. Based on this anonymous information (e.g., sign-ups, browsing behavior, purchase history, and product affinity), potential buyers, known as statistical twins, can be identified on Google or Facebook and contacted via ad campaigns. Again, pseudonymous data ensure privacy in this case. If the potential buyer then enters into a dialogue with the dealer, or responds to the approach by making a purchase, the customer lifecycle is kick-started, and customers are approached in a personalized manner in the future.

Clear is that the role of the marketer will change as a result of the further automation options, namely drifting away from operational execution towards planning and strategy development, and the customer approach will evolve from a one-to-many to a one-to-one communication. This next level of customer communication evolution provides marketers with the ability to target customers and potential buyers, while significantly reducing wastage through ineffective communication and false buying incentives. Customers in turn benefit from suitable offers and tailored special offers.

By Daniela La Marca