The 2016 US presidential election is perhaps the biggest and most public failure of segmentation models in recent memory. Most models not only predicted Hillary Clinton's victory by a comfortable margin but also ignored the demographic, racial, and geographic segments that predicted a different outcome—and propelled Donald Trump's victory.
Taking that failure as an example, and considering the overall outdated mode of assessment that segmentation models seem to be stuck in, we should admit that segmentation models must evolve.
Especially for businesses, an evolution can result in more effective marketing strategies and more successful marketing performance in our increasingly dynamic marketplace.
Why Segmentation Models Could Use an Update
Since the 1980s, companies have spent a lot of time and effort meticulously segmenting audiences on the basis of demographics and creating personas for marketing teams. However, those methods are increasingly irrelevant, for a few reasons.
- First, there is no longer a competitive advantage to traditional segmentation. Every company is doing basic recency, frequency, monetary (RFM) segmentation using the same combination of behavioral or demographic overlay data.
- Second, potential clients change faster than static models. Many segmentation models that are static by design or updated only a few times a year, at best, don't capture consumers' fast-changing behaviors, especially in high-velocity markets such as technology and retail.
- Third, the models' predictive power doesn't translate to actual return on investment. Traditional segmentation approaches were developed for mass-media buying and are not very effective in an increasingly one-to-one addressable world.
Predicting consumer habits solely by focusing on identifiers such as demographics is misguided. Even formerly successful marketing campaigns based on these techniques are now struggling.
Tesco, for example, famously pioneered combining RFM segmentation with one-to-one targeting at scale in the 1990s through its Clubcard program. The program was aggressively digital for its time, rewarding shoppers for buying healthful food and showing loyalty. But because the model no longer provides a competitive advantage, Tesco's 20 million-member loyalty base is not producing growth today.
Benefits of Switching to Needs-Based Segmentation