When one plots the future of Artificial Intelligence (AI) in marketing, one arrives at a singularity where all of marketing is the AI avatar of a brand in direct conversation and interaction with the AI avatar of the consumer. I have called the AI avatar of consumer – Concierge Intelligence in many of my columns here, including my first MxMIndia column back in Jan 2022 -“The Coming Post-Digital Age”.
However, plotting and thinking about the intermediate points would be helpful.
I have been part of a team working since 2020 on using Natural Language Processing (NLP) to generate secondary research semi-autonomously. The launch of GPT-3 and subsequent versions reframed the project for us. Like scores, perhaps hundreds worldwide, we are now trying to find a market niche, proprietary prompt engineering, and the correct interface to support a viable business. Say, a freemium WhatsApp interface for Indian SMEs offering online business consultancy services based on open-source predictive and generative AI models working on public and paid data sets.
What about the emerging role of AI in primary consumer research? The big two—Alphabet and Meta—have been using predictive AI for decades to segment consumers, keep them engaged with their social media feeds and search results, and harvest clicks so their advertisers can pay them big bucks.
Over the past decade, big corporates from both the B2C and B2B worlds have been using Big Data and Predictive Analytics to fine-tune their business and marketing plans. However, it is unclear whether they are at the cutting edge of predictive AI, just as Alphabet and Meta are. While a lot is currently being made of Generative AI and the likes of GPT, Llama, Gemini, etc., I bet that we shall discover that the disruptive power of AI will come not from generating sentences, pictures, videos or music but from underpinning key business, economic, social and personal decisions based on a dynamic array of multi-dimensional data sets. While predictive AI underpins generative AI, a different kind of predictive AI will also underpin the AI age. It will be predictive AI that works on an integrated, dynamic view of the natural world to deliver strategic action plans and monitor and fine-tune them. To use this level of AI, corporations and governments will need to go beyond internal data sets and subscribe to a whole range of third-party data sets.
One category of these third-party data sets will be garnered through an IoT network of sensors synthesised with publicly available identification data sets—for example, vehicle movement with ownership details or scans of browsing shoppers, personal IDs and billing details. The ownership and personal ID can be scrubbed of all details except for basic demographics to meet privacy rules. Alternatively, the individual could opt to belong to an ID Bank that holds his details in escrow and can release them, using blockchain technology on payment of a fee – thus making the individual the valid owner of his ID and personal data.
The ID Bank idea will fuel the second category of third-party data sets. These data sets will contain in-depth profiles of individuals, including contact information, demographics, psychographics, societal and cultural attitudes, media usage, product and brand usage, and purchase behaviour and intentions. The ID Bank will have a watertight agreement with the individual on securely holding the data and releasing any of it to a third party only upon approval and release of a specified fee.
Corporations can then request the release of specified data from a selected consumer profile. For example, a car company may ask for a data set consisting of individuals who own one of a set of car models and have indicated a purchase intention for a new car in the next six months with permission to contact with offers. The ID Bank, in discussion with the consumer, will quote a certain fee on payment, for which the data will be released in a blockchain format that allows for usage tracking. The fee will be released to the consumer’s account, and the ID Bank will get a management fee.
Creating, managing and marketing the two categories of data sets envisaged above will define the future of the market research industry over the next few decades.
The corporation’s predictive AI systems will define the need for data from third-party data sets, consider the cost-benefit of buying them, and incorporate them into predictive analysis to build business and market plans.
Over the decades, as AI and consumers become more sophisticated, intermediaries like ID Banks will be cut out, and a brand’s AI will be in direct touch with a consumer’s Concierge Intelligence (CI) with market research evolving into a version of anthropology focused on studying the behaviour of AI systems. “AInthropology” anyone!?