One of the greatest challenges in marketing is to create personalised communication. Time and time again, it has been shown that relevance drives engagement, which in turn drives an increase in turnover and creates long-term brand value. Of course, the customers know that you have lots of data about them – so this needs to be reflected in the communication. But how do you manage to personalise thousands of engaging customer journeys for customers, when not all of them need to receive an email by Tuesday at 11 AM, and where personalisation by using their names or an individual product category is not enough? This is what the AI-driven customer journey can offer.

Deep dive into data

Data, data, data. This is the current mantra in marketing because data holds the key to relevant and engaging customer communication. The challenge is not only the amount of data, but also the ability to create insight by using this data. That is why companies are looking for technologies that can help them to utilise this potential. One of the more recent members in the marketing technology portfolio is artificial intelligence (AI), that can find the connection within large amounts of data and derive insights that you would struggle to achieve without AI.

When you use AI to personalise your communication, the customer journeys shift from being rule-based to being driven by insights from those data streams that the customers’ interactions with a brand continuously generate. These insights can be felt at more than an aggregated level. AI also helps you to gain an insight into your customers on a more granulated 1:1 level, and it can recommend timing, channel, and content based on subtle patterns.

Strategy is still required

AI is not a quick fix that can vitalise your customer journeys from one day to the next. A clear strategy is required, and you have to know what you wish to achieve with your communication. Which“moments of truth” in the customer journey do you want to impact? My recommendation would be to choose 4-5 scenarios, where you could useAI to optimise them. This could, for example, be the repurchasing of products that are about to run out (replenishment). You can read more about this in the VITA case in the next chapter. It could also be scenarios such as customer retention (anti-churn) or cross sales (share of wallet optimisation).


The first step is to investigate what characterises the customers that match each moment of truth – i.e. the customers that renew their subscription, frequently re-purchase certain products, or who have a high share of wallet. This will give you a clear idea of how to create more of these customers. What are their behaviour and characteristics: Do they look and click on certain links, categories, products? Do they use different services? Do they phone in? Do they redeem points? Have they purchased other specific products, and what kind of people are they in general if one considers classical qualities such as gender, age, demographics?

Besides being interesting knowledge for the company, the patterns also constitute a potential background for an automated scoring of the other customers in the database. Who resembles the customers with the desired behaviour, and to what degree? How can we help them to make the “right” decision during their customer journey, so that they make it to the target?

From a customer perspective, AI provides a more meaningful relationship to a brand during the customer journey, as well as a higher level of relevancy. Increasingly, the communication becomes a service that the customer is interested in, and less of a disturbance that the customer will protect themselves against. Through experience, we know that using AI is rewarded with engagement and loyalty.

Case: AI matches customers and content for VITA

The Norwegian retailer, VITA, that sells cosmetics and skincare products, and is comparable to Matas, has tested Agillic’s AI-driven customer journey.

On average, VITA sent two weekly newsletters with offers that delivered good results. We calculated the average conversion rate and established it as a baseline. Thereafter, we compared it to the conversion rates of the AI-driven communication in a series of scenarios.

Among other things, VITA tested whether the AI-enriched communication could, to a higher degree, motivate the customers to re-purchase products. The customers were, for example, offered shampoo when the probability of re-purchase was high, as the shampoo bottle would probably be empty at that time. The message was clear and unequivocal: VITA was making customers aware of exactly the same products that they had purchased before, and encouraged them to re-purchase these products.

The conversion rate increased significantly, and VITA achieved 10 times better results when using this type of AI-driven communication. This example shows that the AI models can exactly match individual customers with the right product at the correct time – and it can do this for many customers on an ongoing basis. Based on input from the AI, VITA only approached those customers where the probability was high that they would find the shampoo offer relevant.

VITA used the same principle in connection with getting the customers to buy completely new products (share of wallet). Whereas traditional communication without AI was successful in getting one customer to purchase a product, AI-driven communication was able to motivate six times more customers to purchase new products.

The results from VITA illustrate the effect that companies can achieve when using AI-driven customer journeys, as well as relevant and personalised communication.

AI is a breaking point

There is a reinforcing effect when using AI. When you use AI in your communication, you achieve significantly better results, as illustrated by the VITA case. This is, first and foremost, due to the precision and level of personalisation that you can achieve within the communication. Moreover, the effect is strengthened by the fact that the algorithms are self-learning. The AI analyses the results of each communication: what worked well, in which situation did we achieve the sales target with the communication in question, and in which situations did we not achieve the sales target? Through this, the database continuously becomes enriched and “smarter”, meaning that the next communication will be even more precise in terms of personalisation and efficacy.

There is a strong synergy between predictive analytics, AI models, machine learning, and marketing automation. This synergy is paramount in terms of winning and retaining the critical consumer’s engagement and loyalty.

To make it easier to start using all the opportunities that AI offers, Agillic has developed a number of pre-packaged, AI-driven customer journeys. You are welcome to contact me at and +4553 88 65 55 if you want more information about how we can help you get started with AI-driven customer journeys, and what yields you can expect.