Not least the public discussion about ChatGPT has reverberated the topic of artificial intelligence again. More than ever, the public is grappling with the use cases and, not least, the ethical aspects of how to make machines "smarter".
Artificial intelligence is not new
The topic is by no means new; as early as 1956, the term „Artificial Intelligence“ has been shaped. What has definitely changed since then, is the amount of data that can be processed and above all, how fast this is possible. Whereas Apollo 11 was still traveling with a computing power that would have been inferior to any modern cell phone, today we speak of real time as a matter of course. But let's also understand what's behind all this?
It is a matter of course for us that we record and perceive interactions at the moment they happen, process and interpret them, and ultimately draw conclusions and consequences from them. To some extent, the human brain is overwhelmed by this, but machines can do this easily (due to computing power). With the more digitalized world, including the Internet and online services , humans and machines have come much closer togetherin their communication.
But it is not only since this year that attempts are being made to make use of the so-called customer journeys and corresponding touch points, especially in the supreme disciplines of customer contact, marketing and sales. This means asking the question of which path a customer will take in his exchange with the company and where he will come into contact with the company or even be able to make contact. Even 20 years ago, people were aware of this, but did not yet have the ideal tools.
This has changed: Today, the possibilities exist and so do the tools. That's why SAP Emarsys, one of the leading marketing automation solutions, consistently relies on the use of machine-supported real-time analytics. SAP Emarsys calls this real-time interaction with humans in digital touchpoints quite succinctly yet meaningfully "Tactics".
Artificial intelligence offers new possibilities
Tactics, the term actually sums it up quite well. Tactics as a term is to be taken literally in the context of SAP Emarsys. They are ways of proceeding, based on research results and empirical values, how to best achieve certain goals. Here, at the time of this writing, the leading marketing automation offers about 100 tactics for every conceivable recommended course of action for the best possible interaction between companies and their customers at digital touch points - in sales and/or marketing outreach. Through the experiences of the last decades, through the ever more efficient exchange of experiences and through the computing power available to almost everyone, the controls and thus the results of these automations are becoming better and better.
By documentation a Tactic is a predefined automation that has been prepared for a specific marketing/sales application based on the above research, e.g. "acquisition of new customers", but can be customized by the user (Automation Center program). In the process, sectors and industries are also sometimes taken into account and each element of the automation, starting from the start through each condition, is thereby based on empirical findings; seen in this way, currently valid best practice.
It is extremely helpful that on the one hand there are instructions for testing Tactic right on board, and on the other hand that due to the maturity of the automation hardly any adjustments are necessary for customers, so there is little implementation effort and Emarsys - provided the data basis and integration - can be quickly deployed and the advantages effectively used.
Emarsys can be seamlessly integrated
Now, all the algorithms used in Emarsys need data to learn and function. This is where SAP Emarsys comes up with a real power duo and, with a fully integrated customer data cloud, has a perfect basis for playing out and executing personalization as well as segmentation in a targeted manner. It is important to note here: This is not about Un-Supervised Learning or Supervised Learning, therefore a solid data basis is needed (cf. Data Cloud or similar) in order to be able to approach basics for Machine Learning and resulting approaches in the context of marketing automation. Progressive profiling" is used to further enrich prospect profiles and lead profiles and thus improve the results in terms of personalization and segmentation.
At this point, I would also like to point out that it is not only marketing applications and automation suites that make use of intelligent technology. In the sales area, too, decision trees and rule-based decision bases are used to implement intelligent sales management. Intelligent Sales Execution is the name of the add-on and includes automated lead or opportunity evaluation, scoring mechanisms, Next Best Action and other tools that allow sales to become much more effective and efficient. But more on that elsewhere.
Tactics - artificial intelligence implemented graphically
Digital movements can be tracked, for example the scrolling speed on a page, clicks on certain articles, associated keywords that indicate interests. Or even bigger: It is rumored that companies like Meta (cf. Facebook) read out several thousand parameters from users and draw conclusions from them. It is then easy to imagine that these insights can be used to "accompany" the user and thus build an ideal digital experience. It is equally understandable that one can also influence.
How can you make this data useful for your own company? Well, tactics are exactly the tool of choice, since they are, based on our previous description, simplified successive actions and decisions, each of which triggers new actions and in turn reacts to the previous events - this is called automation flow. This is called automation flow. It uses findings from research and experience to ensure that a person or a specific customer is addressed at the right time, in the right context, with the right approach and in the right channel, and ultimately with the right content, in order to achieve a predetermined result (e.g., purchase). But: people are different and peculiar, so every customer journey is also different and peculiar. The resulting hyper-personalization, roughly the claim in offering individuals their own defined customer journey would simply not be possible without automation.
Tactics facilitate the work of users in that at least the individual actions and decisions are predefined and the predictive and prescriptive actions can be derived. In each action step, the tactic determines for each individual customer which follow-up action is to be offered or played out. In this process, the content of web pages is also changed in a targeted manner, certain buttons or links are adapted, expected actions are made more probable, and so on.
More specifically: For example, if I want to attract new customers, based on the experience of customers and researchers, certain actions are optimal when contacting a prospect to have a very high probability of converting them to a customer. Let's imagine that a prospect wants a certain item, even puts it in the shopping cart and abandons it - all detectable by SAP Emarsys. Then it would be a possibility to play a message to this prospect before leaving the browser window, meaning "You are already leaving us, here is a discount coupon". Don't you think? But such still very simple tactics are easily possible.
In an increasingly competitive environment, it is important to use resources, in this case the budget, as effectively and efficiently as possible. This is where key figures such as customer acquisition cost (how much does it cost me to acquire a customer) and customer lifetime value (how much revenue can I usually generate with a typical customer in a certain period of time) come into play. Not to be forgotten is the minimization of bounce rates in order to serve the acquired customer perfectly over the life cycle and thus to bind him to the company.
Approaches to optimize this include customer journey mapping, net promoter score evaluations, service optimization such as One Face to the Customer, and other similar consulting approaches designed to see, understand, and improve the impact of one's business toward the customer.