Spot on Market Research, 3rd Edition

 

Spot On Market Research

April 18, 2018 

MARKET INTELLIGENCE ON MARKET RESEARCH

Ken Sonenclar, Oaklins’ market research specialist, is pleased to share some high-level industry intelligence in this edition of Spot On.

“Our focus in this issue is on artificial intelligence (AI) and the ways it is infusing market research, taking it to places, both exciting and concerning, that could only be dreamed of a few years ago. Before turning our attention to M&A, my colleague John Matthews and I analyzed the commercialization of AI in Europe and North America for New Science Associates, a research and advisory firm that flourished for many years before being acquired by Gartner. As we note in our survey article, the thinking behind many of today’s AI applications was divined decades ago and only needed computing power and memory to become cheap enough to move AI from theory to practice. That moment has arrived.”

ARTIFICIAL INTELLIGENCE AND THE FUTURE OF MARKET RESEARCH

Artificial intelligence promises to revolutionize the field of market research.

Capable of producing an extraordinary wealth of timely and market-relevant insights, AI should also dramatically reduce the cost of gathering data and producing market research reports.

Newer market research firms will use AI to win business from established market research firms (becoming acquisition targets in the process). To stay competitive, older companies will need to adapt, integrating new technical skills to utilize AI in data gathering and analysis, such as that collected from social media and smart sensors, and automating report production. 

However, there are obstacles ahead for the application of AI in market research given the tightening of data privacy regulations in some parts of the world, as well as the sometimes justified and sometimes irrational fear of AI shown by many consumers.

Students of artificial intelligence have long recognized that the cost of computing power — which AI consumes like an AMG-class Mercedes — has long impeded its commercialization, but processing capability has increased to the point where AI has many useful applications. The plunging costs and increased capacity of data storage, along with the sheer volume of useful, real-world data to work with, drive the need for new and novel methods such as AI to filter, categorize and make sense of this deluge.

Machine learning (ML), with roots in 1950s defense research, is now widely used to drastically reduce the time needed to produce a new data analysis or process optimization model. Such models are generated by entering vast amounts of before-and-after case data into an ML algorithm. This can be useful in flyby-wire process optimization for running marketing campaigns and producing predicative data analysis models. It is also used to improve natural language processing (NLP) and computer vision recognition of objects and faces (discussed below). Machine learning underpins most other examples of AI as well as being an AI technology in its own right.

Download the full newsletter from the tab at the right