• John Ennis

Eye on AI - August 29th, 2019


Welcome to Aigora's "Eye on AI" series, where we round up exciting news at the intersection of consumer science and artificial intelligence!

Today’s theme is the growing trend of using publically available datasets to gain new product insights. While I personally believe this approach is flawed and will lead to negative business outcomes unless used carefully, there’s no doubting its popularity and the importance of being aware of trends derived from publicly available information.


Look for more from me on this topic in the near future. But for now, let’s review this week’s news!


Old Companies Increasing Reliance on Digital Data

We begin with an apt article from the Financial Times, titled “Big brands turn to big data to rekindle growth” (trial subscription may be required to view for non-FT subscribers, but worth it for such a relevant topic), which describes how large, established consumer companies like PepsiCo, McDonald’s and Danone have begun to rely on AI-based market research firms to help them predict customer trends.

“Black Swan [one such AI market research firm]… hoovers up data from social media, online forums, product review websites as well as other sources and then analyses it to divine what consumers want. Its artificial intelligence software purports to sift signal from noise to figure out which early trends are destined for mass adoption...”

Many of the brands adding market-research-AI practically invented consumer research when it first emerged, but have now turned to new methods as influence and profits have been declining since 2012. The problem, according to Tim Warner of PepsiCo, is that they’re old methods aren’t quick enough to take advantage of emerging trends.


“The old methods that were invented before the digital era are not agile, precise, and predictive enough,” adds PepsiCo’s Tim Warner. “Leading consumer goods companies want to upgrade decades-old techniques, such as consumer surveys, focus groups and retail sales data, which are seen as too slow, too expensive and often incomplete.”

Many companies have even turned to digital data as the end-all of market research, replacing traditional market research methods (and workers) with Amazon reviews and social media posts.


“... instead of commissioning an ‘in-home use test’ that follows 400 people using its Cif kitchen cleaner and a competitor product,” writes Stan Sthanunathan, leader of consumer insights at Unilever since 2013; “Unilever can now just mine product reviews on Amazon to compare Cif’s product performance and that of its peers.”

Will classical consumer research be replaced altogether by mining publically available datasets? Time will tell, but I believe that those who abandon classical consumer research will come to regret their decisions. One reason for my belief, out of many, is our next topic.


Suspect Amazon Reviews Point to Corruptibility of Digital Data



But there’s an inherent danger in any market research that relies solely on digital data, which is that it’s often flawed, as outlined in MediaPost’s article titled “Amazon Still Dealing With Fake, Hijacked Reviews.”

“Fakespot, a New York-based startup [that identifies fraudulent reviews], estimates that overall 34.65% of Amazon reviews are suspect, according to one report,” writes author Laurie Sullivan. “-- and the problem only continues to worsen as the company grows.”

Digital systems that track online data are often exploitable, leading to misleading reviews and other insights. I’ll be talking about this topic in much more detail in a blogpost soon, so stay tuned!


Other news:



That's it for now. If you'd like to receive email updates from Aigora, including weekly video recaps of our blog activity, click on the button below to join our email list. Thanks for stopping by!


©2020 Aigora