AI Leads to More Effective Marketing Activities
Consumers have high expectations regarding availability, usability, customization, interaction and content. At the same time, digitalization has caused an explosion in the volume of data that can be collected at the various digital touchpoints along the customer journey. These two developments boost the use of marketing technologies that help marketers analyze data and draw the correct conclusions for effective marketing activities. More and more of these applications also use machine learning models. They help marketers interpret the data generated along the customer journey better, faster and more precisely.
Apart from “AI first” applications, which provide machine and deep learning as core services, the major MarTechSuites also increasingly include AI components such as Cortex in Sitecore, Einstein in Salesforce or Sensei in Adobe. Other platforms such as Shopify or Hubspot have established marketplaces where marketers can purchase AI services. AI is increasingly becoming a standard fixture in the marketing world.
AI Along the Customer Journey
Artificial intelligence can be used at the various touchpoints along the customer journey. The implementation requires consideration of the individual touchpoint as well as the journey in its entirety. Amazon optimizes the user experience at every touchpoint using machine learning: With programmatic sampling, Amazon introduces visitors to new and potentially interesting products at an early stage in the customer journey.
Machine learning supports the selection of products displayed. As a market place, Amazon offers its sellers precisely this product placement, for instance, by offering a free sample. In combination with re-targeting ads, this leads to a 15 percent increase in sales conversion on average.
AI: Automate, Optimize, and Augment
The main goals of machine learning in the context of customer acquisition and retention are automation, optimization and augmentation.
AI can be used to render existing processes as efficiently as possible. This is why many companies are investing in AI. According to Gartner, in 2020, 25 percent of all customer interactions with a company will take place without human involvement. Commerce and assistants are typical use cases for this. The optimization of direct interaction between human and machine is considerably more complex. The data and models involved are being improved continually. Natural language processing is an excellent example of this.
The greatest potential of artificial intelligence lies in augmentation: Here, human and machine work hand in hand. The marketers take care of all tasks requiring empathy and creative solutions. The machines do the groundwork and help the marketers make better decisions. One example are the AI features of Salesforce Einstein that suggest customer-specific solutions such as free shipping or interest-free deferred payments. In the end, however, the agent decides which offer fits the customer best.
Marketers’ key areas of interest in machine learning can be clustered into three types: predicting behaviors, anticipating needs, and hyper-personalizing messages. The overarching goal of machine learning in marketing is personal, tailored and seamless interaction with the consumer.
The User as a Driver of Personalization
The process of personalization is an infinite loop comprising three steps: A system action triggers a user interaction, which generates data, which in turn triggers the next action. Data, actions and feedback options need to be prepared upfront. As long as the user remains in this infinite loop, data and therefore marketing intelligence is being generated.
The challenge lies not in the volume but also in the quality of the data. Marketers have been capturing data for years, but only machine learning has made it possible to make targeted use of this. However, many marketers underestimate the time and effort required to process the data and improve quality. This leads to the termination of many AI projects. In many cases, the data collected includes information that is completely irrelevant. As they say, “garbage in, garbage out.” It is becoming increasingly important to understand the data in order to improve the machine learning model and generate relevant results. One of the big questions in marketing is whether users are willing to give up their data for better personalization.
When data is available, the next question is which action to recommend to the user based on the data.
For this, it is not enough to just understand the data – it requires a much more comprehensive view: What is it we “sell”? What is the best action to take next? User interaction is the driver behind the entire process. Users expect intelligent and appropriate actions for them to react to. The better the actions, the more valuable the interactions. Understanding and controlling this process is a challenging task.
Machine learning further increases the complexity of the process. Increasing customer loyalty is an ambitious goal that requires a lot of time and resources.
AI Requires a New Mindset
Artificial intelligence is not just another marketing technology. It turns the way we do marketing on its head. Machine learning enables seamless, personalized user experience. Innovative marketers are already exploiting the potential and are learning how to improve.
In general, machine learning leads to a higher degree of automation, optimization and therefore more efficient and more effective market cultivation.
Machine learning promises to make marketing more efficient and at the same time more human. Artificial intelligence can support every functional area of marketing and every step of the customer journey by linking data to actions in an infinite loop. The key question is where humans are indispensable in customer care and where machines can provide better service. The challenge lies in finding the right balance. To do that, you need to keep in mind the fact that artificial intelligence can only create added value when it is based on a solid foundation of technology, data and processes as well as a strong organization with innovative employees, interdisciplinary skills a culture of responsibility. It is not about algorithms but about understanding the customer journey and user experience. This is, after all, an “old” core discipline in marketing.
This contribution is based on the research report by Alex Mari “The Rise of Machine Learning in Marketing: Goal, Process, and Benefit of AI-Driven Marketing”, University of Zurich. The study was published in May 2019 following the open research principle.
Study design: 32 Swiss and international experts and executives from the fields of marketing, IT and Data were interviewed. The interviews were transcribed, coded line by line, grouped and structured, giving a network of 20 main nodes and 76 sub-nodes. The results are depicted in the conceptual framework of the AI-driven marketing model.
- Gartner “8 Top Findings in Gartner CMO Spend Survey,“ Chris Pemberton, November 5, 2018
- ChiefMarTec “Marketing Technology Landscape Supergraphic (2019),“ Scott Brinker, April 4, 2019
- Medium “A Map of Amazon and Modern Marketing,“ David J. Carr, September 11, 2018