How is Artificial Intelligence being used in Marketing, and more generally in end-customer relationship processes, i.e. Sales and Customer Service?
Evidence from the latest census conducted by the Milan Polytechnic Observatories shows that, to date, most of the projects related to the use of AI algorithms in companies concern the areas of customer service operated through virtual assistants and chatbots. In particular, chatbots are used by as many as 81 percent of organizations and are, therefore, quite widespread as are voice assistants (83 percent). There is, however, growing interest in eCommerce recommendation systems due to their demonstrated effectiveness “in the field” – one in four users, according to respondents, have finalized a new online purchase after receiving targeted advice.
What is Artificial Intelligence in Marketing
In general, Artificial Intelligence is a set of hardware and software systems endowed with typical human capabilities (interaction with the environment, learning and adaptation, reasoning and planning), capable of autonomously pursuing a defined purpose by making decisions that, until then, were usually to given to humans.
Instead, Artificial Intelligence Marketing (AI Marketing) is called Marketing that uses Artificial Intelligence to interact with customers, improve understanding of the market and people, and suggest-more quickly than humans-actions to be taken to refine persuasion techniques.
Artificial Intelligence in Marketing leverages the latest technologies that fall under the umbrella of AI, such as Machine Learning and Nlp – Natural Language Processing, integrated with mathematical/statistical techniques (such as those of Bayesian networks, a probabilistic graphical model representing a set of variables with their conditional dependencies) and behavioral Marketing (behavioral targeting). All with a very clear and direct objective: to improve persuasiveness in order to lead users to “convert” the company’s “call to action,” i.e., to take an action that generates value for the user himself but also has a positive implication for the company.
From the cycle of “perception-reasoning-action” to “collection-reasoning-action”
In other words, AIM provides CMOs (Chief Marketing Officers) with a set of tools and techniques to guide the behavior of target users, those a company intends to address.
The principle on which this new branch of Marketing is based echoes the “perception-reasoning-action” cycle typical of the cognitive sciences, which in the context of Marketing becomes “gathering-reasoning-action.”
1) Collection
The first pillar of the cycle refers to all those activities aimed at capturing data from customers, potential customers and, more generally, people “on target” with respect to the goals of the company or a Marketing campaign.
2) Reasoning
This is the part where data is transformed into information and finally into intelligence or insight, the central part where Machine Learning and Artificial Intelligence play a central technological role.
3) Action
The intelligence and knowledge achieved through the reasoning phase are the ones that then allow action to be taken; in the context of Marketing, action can be translated into a communication or campaign with a higher probability of persuading target users (and thus with superior results in terms of effectiveness for the company).
A cycle that, as a whole, could actually be fully automated through a more widespread use of Artificial Intelligence technologies in all phases of the cycle, including that of the actions to be deployed.
The benefits of AI applied to marketing and customer experience
Adopting an approach that integrates artificial intelligence into the marketing funnel and the user’s Customer Experience, therefore, would have a strong impact on the effectiveness of the strategies implemented.
The power of new enabling technologies, applied to marketing, comes in the form of 4 competitive advantages, on multiple levels and in a multiple domains of action:
- AI is essential in the data collection phase, as by integrating with major marketing campaign management platforms and CRMs and tools for mapping customer interactions, including from a sales and post-sales perspective, data can be extracted in a timely manner. Not only that, the resulting data collection is able to generate reports that highlight patterns, new opportunities and true actionable insights.
- Predictive models become possible, just from the data-lake mentioned in the previous point. Fed to an intelligent algorithm, all details about customer behavior can be translated into predictive marketing actions, intercepting new interests and needs even before they become apparent.
- ROI is maximized and diversified as it becomes possible to analyze actions in real time, adjusting the focus and declining campaigns toward new opportunities and market targets.
- The integration of human intelligence with artificial intelligence yields extraordinary results. One should not think, as is the fear of many, that AI will replace the human factor. There is a two-way marriage between the two intelligences, which is necessary and mutually positive, because on the one hand algorithmic intelligence has a very high degree of accuracy and objectivity — not being influenced by bias and opinion — but on the other hand the human element has a figure of creativity and non-linearity that are essential for the success of marketing activities.
Areas of application
Aggregation and analysis of data(including unstructured and natural language-based data) in a continuous process of learning and improvement to identify from time to time the most probabilistically effective communication and sales actions, strategies and techniques (those that have the highest potential for effectiveness/success for individual user targets). This is, in essence, what AIM does.
Thus, starting with this sort of identification of the technological scope of Artificial Intelligence Marketing, we can recognize the specific target applications and technologies and how they can be employed.
The most popular applications in marketing are as follows.
Virtual Assistants and Chatbots
These are software that can perform actions or deliver services to people based on commands or requests received through natural language interaction (written or spoken). The most advanced systems are capable of understanding tone and context of the dialogue, storing and reusing the information gathered, and demonstrating resourcefulness in the course of the conversation. These systems are increasingly used as the first level of customer contact in the for assistance through corporate customer care.
Recommendation Systems
Widely used, for example, in eCommerce or video and music services, they are solutions geared toward targeting the user’s preferences, interests, or more generally decisions, based on information provided by the user, either indirectly or directly. In essence, they are personalized recommendations about one’s tastes, for example, based on previous purchases, which may be located at different points in the customer journey or, more generally, in the decision-making process.
Technologies
Looking instead at technologies, one classification is as follows.
Content creation and curation
The automatic creation of content (articles, news but also “simple” messages) and its presentation to the correct audience, at the optimal time (i.e., the one where there is the highest chance of persuasion and, therefore, call-to-action conversion), is one of the most promising areas of Artificial Intelligence Marketing.
This is where advanced data analytics systems, event correlation, natural language understanding, analysis and recognition of images, video, voice, and, across the board, self-learning techniques (based on Machine Learning systems and algorithms) come into play that allow the systems that create and propose content (reading recommendations, related images, personalized ads, etc.) to continuously improve the proposal capability; all dynamically depending on how people actually enjoy that content and the potential of each in terms of persuasion and conversion.
Voice search
Voice search is one of those technologies that has now entered collective public appreciation thanks to systems such as Apple’s Siri, Microsoft’s Cortana, Amazon’s Alexa or Google’s Google Now. What CMOs need to know is that these technologies are changing SEO – Search Engine Optimization techniques and strategies and that voice will be one of the most impactful elements on organic content traffic.
From a Marketing perspective, this trend should be “ridden” by leveraging chatbots and AI-based virtual assistants to “guide,” advise and persuade users.
Programmatic Advertising
Machine Learning algorithms represent the technological basis through which to model and analyze the purchase (or action) propensities of target people in order to distribute advertisements and communication ads in a more targeted way.
Not only that, it is also through Machine Learning that companies will be able to have a closer control over the purchase and distribution of their ads on automated platforms such as those, for example, of Google (by way of example, identifying more accurately the sites and the types of users who browse on them in order to verify their trustworthiness or alignment with the positioning, strategy, and reputation of their company).
Target and propensity modeling
The success of a Marketing campaign, strategy or action depends first of all on the correct identification of the target audience and the analysis of the propensity of the people in the target audience to perform a certain action (object of the Marketing proposal). And it is perhaps on these aspects that Artificial Intelligence Marketing succeeds to its fullest business potential.
Machine Learning algorithms in this case represent a keystone from the past because they enable a process of continuous improvement (based on the analysis of big data and persistent learning) in times and with precision unimaginable for humans.
Propensity modeling then opens the door to additional analytics specific to Marketing actions, such as real-time pricing and rating of activities with the highest probability of success. Activities that then result in Marketing actions/activities that are faster, cheaper for the internal organization and more effective for the business.
Marketing Automation
Marketing Automation usually encompasses a set of (automated) rules and activities that serve marketers and CMOs to manage and optimize demand generation, i.e., the process of acquiring and managing prospects (until they “transition” into actual customers), of which lead generation (acquisition of prospects), lead nurturing (curation and management of these prospects) and sales conversion (the transformation of these users into actual customers for the company) are part.
Again, the technological basis of reference, remaining in the area of Artificial Intelligence, is Machine Learning through which to analyze all user data (from any touchpoint and channel) and estimate the most suitable lead generation and nurturing activities with the highest probability, again, of being translated into effective sales conversations.
