skip to Main Content
How to Read Your Customers’ Minds

How to Read Your Customers’ Minds

Mind-reading used to be a popular and profitable con back in the day.  The so-called “Mind Reader” would arrange to have a shill in his audience. When the Mind Reader called on this paid plant, he would miraculously know all about this person — thus “proving” his telepathic powers. Well, the art of mind-reading is back. And this time, thanks to the incredible capabilities of Artificial Intelligence (AI) and Machine Learning, it is definitely no con.

Case in point: Levy Restaurants in Chicago, which offers vending and food services to major event venues, wanted to appeal to the changing tastes of their target customers, the city’s sports fans. They discovered through an AI-powered analysis of social media activity that those customers were interested in more fusion cuisine. Knowing this was the case, they changed up their restaurant concepts for the upcoming season.

Result? Within half of that season, their same locations made more in revenue than they earned in the entire previous year.

Whether you’re a B2C or B2B company,  there is nothing more important than understanding what your customers want, and AI and ML can help.  In the words of Casey Carey, Senior Director of Product Marketing for Marketo Digital Experience at Adobe, “AI will become pervasive throughout B2B marketing efforts, improving performance and increasing efficiency throughout the entire buyer’s journey.”  

AI and machine learning also provide a distinct advantage when it comes to mining insights from market research. For example, a Northern Light client wanted to know if machine learning could help them figure out what hashtags were relevant to a particular market segment.  There were thousands to evaluate and new ones were being created every day. Northern Light created a machine-learning powered solution that mines Twitter for sample data based on new hashtags and semantically evaluates the content for suitability against a training model that represented the client’s market segments.  For those that pass the suitability test, Northern Light then adds all the content using those hashtags to a database where customer attitudes can be studied by the market research team and analyzed 100 ways from Sunday. Insights started to emerge almost immediately as the new system ran because hashtags are self-applied labels by customers.  

I think this example illustrates how to use machine learning to gain insights into your customers:

1)   Focus on an important and immediate problem that can’t be solved any other way.

In the above example, the customer had a pressing need to identify relevant new hashtags and the traditional method of manually eyeballing a Twitter search result was not adequate from speed, cost, and scalability perspectives.  Solving that problem with smart automated means gave the market research team what it needed. And there was no other way to do it than letting the machine reason through the thousands of candidate hashtags.

2)     Solve the whole problem, not just the machine learning part.

Many machine learning projects fail out of the box because it is too hard or expensive to accumulate the content needed to train the models.  So work to get the content first. And make sure the project includes all the parts that make the solution a complete system. In the case of the hashtag analysis project, that included using the output of the machine learning algorithms to automatically accumulate the research database of tweets for later analysis using other tools.

3)   Use machine learning where it can add value .

  An executive at a company I talked to once about becoming a client made what I nominate for “Clueless Comment of the Year.”  He said, “Can’t we just buy one of those AI programs and turn it loose? If it’s so smart, want it figure out what it needs to do?”  The short answer is “no, it won’t.” Machine learning can add value in very specific ways and solve very specific problems, It is not generalized human-like intelligence.  So focusing on problems that can be solved today by the technology we have is key to successful AI projects.

Fully understanding what your customers want has always been the key to effective marketing. Now, thanks to AI and machine learning, actual telepathic abilities are no longer required to gain that incredibly valuable knowledge.

Back To Top