The latest MediaPost article on the K-shaped advertising economy in the United States by Joe Mandese caught my attention because it isn't really about advertising.
It's about something much bigger: heterogeneity.
The article shows that while overall ad spending grew modestly, some sectors accelerated while many others declined. The average describes almost nothing because very different realities are happening at the same time.
That's exactly the problem with many of today's marketing measurement models.
Most still estimate an average elasticity, average conversion rate, average lift, or average ROAS. But those are often nothing more than "averages of averages" built across customer segments, retailers, markets, products, and time periods.
When customer groups are moving in different directions, averaging them together doesn't simplify reality. It hides it.
One segment may be responding to pricing. Another to merchandising. Another to assortment. Another to media. Some customers are buying more frequently while others are reducing trips. Some are trading up while others are trading down.
Those aren't exceptions. They are the commercial system. At Provalytics, we believe growth comes from understanding that system rather than compressing it into a single response metric.
Our Agile Insights Modeling (AIM) connects multiple marketing investments, multiple customer responses, and multiple business outcomes simultaneously because that's how businesses actually operate.
The K-shaped economy isn't exposing weaknesses in advertising. It's exposing weaknesses in models that assume everyone behaves like the average customer.
Customers don't respond as averages ... neither should our models.
Cheers
Todd