AI and Marketing Mix Modeling

In the course of my research on the impact of artificial intelligence on sales, marketing and customer experiences, I have been learning about Marketing Mix Modeling (MMM).  If you are not familiar with it, here is a quick description from our friends at Wikipedia.
"Marketing mix modeling (MMM) is a way to optimize advertising mix (where money is spent on advertising) and promotional tactics with respect to sales revenue or profit.  It is an analytical approach that uses historic information, such as syndicated point-of-sale data and companies’ internal data, to quantify the sales impact of various marketing activities." 
As I dug deeper into MMM I saw both the value and the complexity involved.  However, my liberal arts degree, in no way prepared me for it.

The ideas and concepts around MMM have been used in the CPG (consumer packaged goods) industry for decades, but little outside of that industry.  The reason - it required massive volumes of data and greater computing power than was available at an affordable price.  Today, however, with digital transformation, increasing numbers of digital customer interactions, the abundance of data, algorithms, analytics dashboards, artificial intelligence and everything-as-a-service, MMM is rapidly expanding.


I recently attended a conference session presented by Amit Satsangi, a data scientist with Oakwood Worldwide.  In his research he has identified seven observable patterns in people's responses to different forms of advertising:
  1. Current Effect
  2. Carryover Effect
  3. Shape Effect
  4. Competition Effects
  5. Dynamic Effects
  6. Content Effects
  7. Media Effects
Each advertising format and its effects - are measured and assigned values using mathematical formulas.  Each "effect" is measured, but not in isolation.  Any of the seven "effects" can change in value when grouped with other forms of advertising.  As you can see, the complexity exponentially increases when different advertising formats are combined into campaigns.

The value of MMM is that through the use of mathematical formulas, data collection and algorithms to analyze the data from past "effects" of advertising, you can better plan where to spend your marketing dollars in the future to maximize sales and the ROI.

So when trying to determine where to spend marketing dollars, Satsangi would advocate measuring each form of advertising by their seven effects (using complex math) and giving each form of advertising a score.
  • Paid Search (27)
  • Print (5)
  • Research Reports (30)
  • Events (10)
  • Social Media (28)
In this example, for every $100 of spend, the formula would suggest you spend $27 on paid search, $30 on publishing research reports and $28 on social media advertising.  This example is not suggesting you do not spend on print and events, it is suggesting you spend only the amounts they deserve.

Now really who can do this on paper and pencil?  None of us, but when the data capture is automated, and the formulas and algorithms get embedded in artificial intelligence systems so marketers can simply view a marketing strategy and spend dashboard, then the adopters of MMM will leave all others behind.

************************************************************************
Kevin Benedict
Senior Vice President Solutions Strategy, Regalix Inc.
Website Regalix Inc.
View my profile on LinkedIn
Follow me on Twitter @krbenedict
Subscribe to Kevin's YouTube Channel
Join the Linkedin Group Digital Intelligence
Join the Google+ Community Mobile Enterprise Strategies

***Full Disclosure: These are my personal opinions. No company is silly enough to claim them. I work with and have worked with many of the companies mentioned in my articles.

Interviews with Kevin Benedict