Cognitive marketing: ‘Packaged intelligence’ in action

Cognitive marketing’ is coming of age with the emergence and increasing adoption of Artificial Intelligence driven marketing applications that can think and act on their own.

Machines are increasingly going to take more marketing decisions. It may appear far-fetched, but it is definitely possible now with the advancements in Machine Learning and Artificial Intelligence domain. Some call this ‘cognitive marketing’, which will signal the emergence of many marketing applications which can think and act by itself. And some call it marketing automation 2.0 to refer to the mass adoption of artificial intelligence driven automation of digital customer engagement activities.

The drivers

Machines are becoming ‘smarter,’ thanks to the advancements in computing, sophistication in analytics technologies and access to large volumes of data to create broader and deeper intelligence. However machines are not only good at decision making, but also in acting on them seamlessly for digital customer engagement making use of smart devices, sensors and touch points.

Chatbots are seen as an early form of such self-learning intelligence being put into digital marketing commercially. They come with built in intelligence for driving contextually relevant engagement with customers, again saving effort and resources of customer service team.

Cognitive Intelligence driven Marketing

With ‘smarter’ machines, marketers can look forward to focus on strategic decision making leaving operational decision making to machines. A cognitive marketing application can leverage Artificial Intelligence to independently make decisions learning from large volume of data sets. This will increase the speed of decisioning which will in turn improve operational efficiency. Subjective heuristic manual rules will be replaced by machine learning driven cognitive decisioning.

Cognitive intelligence is achieved by ‘packaging’ various analytical models that continuously infer intelligence from underlying data and then embedding them to the decisioning workflows. For example a typical customer lifecycle marketing program can be automated with the application of various ‘packaged analytical models’ across the workflow as shown below.

Since operational decisions are many and often repeatable in nature, machines can take the role of marketers in operational decision making and execution. Since Artificial Intelligence emphasises on feedback based optimisation, the operational decision making gets better over time just like how human make better decision with more experience and market know how. However machines have a significant edge over humans in the sense that decisions are always made based on facts and subjective biases doesn’t influence decision making, again adding to the efficiency.

Man Vs Machines

It is a widely accepted notion that humans can do a better job when it comes to taking decisions that need emotional and innovative thinking. However, the definition of ‘smartness’ has changed over the years and is now typically associated with academic performance or efficiency of doing tasks with minimal errors. This is where machines are all set to compete with humans and even outsmart them.

Smart machines can ingest, process, store, and access information faster and efficiently than humans. Right kind of data can be ingested and prepared for analysis, this is what I call ‘smart’ and ‘good’ data which has more intrinsic value. Artificial Intelligence and machine learning can derive patterns much faster and produce a wider choice of options and predicted outcomes for making decisions better suited to a given problem and context. With the latest deep learning techniques using feedback loops, machines can even learn faster optimizing the decisions and actions continuously.Examples of Assistive Intelligence are automated marketing campaigns or automated back-office functions such billing, customer service and so on. The machine learning capability of the assisted intelligence allows for quality improvement of efficiency matrices.

Assisted intelligence is all about empowering machines to execute actions. Objective here is more about improving operational efficiency, through clearly defined rule based repeatable tasks that can be automated using a software application or a physical robot, simulating the activities of human beings.

Augmented intelligence is all about man-machine collaborated decision making. Today machines can make use of advanced analytics and visualisation capabilities to come up with deeper insights as well as precise recommendations for humans to take decisions in the moment of relevance. Suggesting a recommended list of offers for a customer service agent to extend to customers walking by to the store is a good example.

Autonomous intelligence is about machines taking decisions and acting upon them without direct human involvement; it is the promise of future. There are some early adaptors, for eg: today NASDAQ has automated 75% of its trading using autonomous intelligence. Other use of this, would be in self-driving cars.

With the advancements that is happening around AI, self-driven products and services with in-built intelligence, conversational and marketing capabilities is not a distant dream. The only challenge may be ability to adopt to changing technologies and incorporating it to build complex, but user centric applications for the enterprise to fully automate their digital customer engagement activities.

What does a typical customer purchase journey look like? A customer sees your product, buys it then may repeat the purchase if they are satisfied with the outcome.