April 2019, Brent J. Dreyer, Managing Partner, DataEM.com
We can all agree that autonomous "self-driving" vehicles have been at the forefront of media over the last decade, but what about autonomous marketing?
The verb autonomous is defined as something having the freedom to govern itself or control its own affairs. That is pretty scary in a car and even scarier as the concept emerges into all channels of our society. Yes, that will include marketing.
In today's world, people do not want to research data, wade through analytics or even depend on other people to do that for them. Basically, people would like the key performance indicators presented to them with the immediate option of taking action. From a marketer's perspective, taking action may mean to isolate and correct issues, create new marketing segments based upon observed or predicted behavior, pause/start campaigns, all from a single interface.
Dashboards might look pretty, but most of them are not actionable,
and therein lies the problem with 99% of the dashboards that are out there.
In contrast, the futuristic automotive dashboard with a marketing interface (portrayed above) displays the key performance indicators and provides the option of doing something about it from the same interface.
Dashboards in the future will all be actionable, and more appropriately, they will be called control panels. With Artificial Intelligence and machine learning being implemented to the fullest extent, you will see issues and opportunities and be able to ask for more information and even resolve problems.
ENTER THE REALM OF
THE MARKETING VIRTUAL ASSISTANT
In November of 2014, Amazon released its Amazon Echo, aka Alexa. At this point, the voice-activated virtual assistant was injected into society. In January of 2019, Amazon announced that it had sold over 100 million Alexa-enabled devices.
What about a marketing virtual assistant? Well, Amazon is actively promoting the use of Alexa in Enterprise and Business Solutions, and slowly, quietly, privately, these uses are coming online. Providing that you have your data in order, it is not as difficult as you might think to create an interactive assistant. Amazon offers a strong set of technical support for building your own skill set.
Imagine the following dialog with your Amazon Alexa or Google Assistant. Ten years ago, this type of interaction would have seemed like science fiction. Today, this type of interaction is not only believable, it is doable and being done.
It's Happening Now
There are platforms and systems in place today that start to mimic parts of Autonomous Marketing, providing a hint of what we can expect to be integrated in the near future.
Marketing Automation helps marketers become more efficient. As illustrated in the Google Search Trends graph, the term Marketing Automation began gaining popularity around 2011.
Marketing Automation helps marketers plan, coordinate, measure and manage their marketing campaigns. Most of the systems set up routines that trigger an action. This includes functions of Lead Generation, List Selection and Segmentation, Scoring, and building Predictive Analytical Models. Many of the CRM (Customer Relationship Management) systems are also included. Basically, marketing automation systems know who to market to, when to market to them, what to say and which marketing channel(s) to use.
Customer Data Platform, also known as a CDP, is a term that was first coined in 2013 by David Raab, founder of the CDP Institute. As illustrated in the Google Search Trends graph, the term Consumer Data Platform began gaining popularity around 2016.
As defined by the CDP Institute, "A Customer Data Platform is packaged software that creates a persistent, unified customer database that is accessible to other systems".
We have explained the concept of a CDP with the analogy of a candy wrapper. From left to right, data is collected and aggregated from multiple sources. The first twist in the wrapper is where the ETL (Extract, Transform, Load) occurs, making the data all look alike.
2019 Copyright DataEM.com
The magic happens in the candy where the CDP unifies the data, pulling together individual identities, scoring and building segments of matched behavior. At the next twist, the appropriate collection of individuals are selected by business rules. Then they are pushed at the right time, through the right channels, with the right message, to the right people.
This is similar to the marketing automation systems, however, their triggered events typically have more vertical functions whereas a CDP will initiate a broader set of marketing events.
Finally, the individual's actions and responses are collected from the marketing channels and fed back into the data sources, as shown in the data flow arrows. The process starts again and the system continues to learn and grow in functionality.
As you can surmise, we are getting very close to seeing autonomous marketing systems. Perhaps the first will be marketed by Ron Popeil with the tagline, Set It, and Forget It! (for those of you old enough to remember his infomercials).
As inferred in the definition, an autonomous marketing system needs to have the freedom to govern itself or control its own affairs. To do this, our current marketing systems will need more intelligence and that will come in the form of AI (Artificial Intelligence).
Today, a marketer looks at analytics and key performance indicators to see what is working. In the future, smart systems will make those evaluations, automatically adjusting spend, content, channels, and personalization in marketing communications.
AI, or machine learning, is already applied within a number of CDPs today. Algothrims are predicting behavior, scoring segment, identifying sentiment, and triggering marketing campaigns. In fact, the ad networks are using algorithms to budget and allocate ad spend, intelligently delivering content and banner ads.
Machine learning is becoming commoditized, making it very easy to incorporate. Using the intuitive interface of our machine learning partner BigML, in a matter of hours, we created an unsupervised learning model that looks at website visits and predicts clusters of buying patterns, identifying the time between visits and what will be purchased next.
The Technology Is Here
Autonomous vehicles are heavily dependent upon AI, constantly learning through feedback from CNNs (convolutional neural networks). Deep learning models like CNNs are very transportable to other modalities and vertical applications, such as marketing systems. Marketing Automation and CDPs hold the mechanical pieces to the autonomous puzzle and AI Democratization will provide the sharing of intelligence between systems. With all of this in mind, you will soon see Autonomous Marketing Systems entering the marketplace.
2028 AMS CONTROL PANEL
(Autonomous Marketing System)
We expect to see AMS (Autonomous Marketing Systems) products fully functional within the next 10 years, by 2028. Hence, the "Dashboard" as we know it today, will be replaced by Control Panels. The big distinction between Control Panels and Dashboards is that Control Panels are actionable. Similar to the dashboard, you will be able to see what is going on, whereas with an AMS you will be able to verbally ask for more information, ask for recommendations, activate and regulate channels, start new tasks (campaigns), manage budgets and set warning flags.
If you would like to explore and experiment with the design, development, and build of an AMS Prototype platform, we have the knowledge and the right partners to pull the pieces together. Contact us directly to discuss the requirements, Brent J. Dreyer, Managing Partner, DataEM, BrentJDreyer@DataEM.com
Autonomous Marketing will employ Automation Layering between the Marketer’s interface and the Artificial Intelligence powering the Autonomous Marketing system. This will allow for control of a platform that can get out of hand, providing results beyond the original design, or missing opportunities that the AI system was not designed for.
To accomplish this, Automation Laying has four requirements for development.
Monitoring and Modifications
System Training involves supplying the system with the appropriate data so that the AI algorithms may begin to recognize new patterns. From this, new rules are inserted to learn a desired outcome.
The Automation Layer must be monitored to identify unwanted anomalies, warranting modifications that will correct its output that guides the AI system.
While the Automation Layering will enhance the underlying AI Marketing System, it needs to be managed to ensure those desired outcomes. For example, Autonomous Marketing systems typically have an allocation of funds based on a budget. As opportunities rise and fade, the budget may be increased or decreased beyond the original thresholds.
AI Marketing Systems will not find new market opportunities without a Marketer pointing them out. This is where the role of an Opportunity Strategist becomes important. Before new rules are injected into the Automation Layer, their viability and sensitivity must be tested and put through a business valuation. Today, this is accomplished with A/B testing of outcomes but in the future the comparisons will be more complex.