Marketing researcher Kristi Price asks us to imagine a world where every ad we see is relevant. Currently, ads are not very well targeted. Marketers aspire to deliver the right ad to the right person at the right time, but clearly, they are missing the mark. Price relates that she sees ads for a pair of shoes that she just bought. She believes the advertising industry can fix this with better marketing math. According to Price, marketing analysis can produce much better results if it involves more quantitative marketing research.
Price started out in finance. When we compare the evolution of finance and marketing during the digital revolution, we see some differences. In both cases, the digital revolution makes it possible to look at vast amounts of data in real-time. The finance industry has embraced this opportunity and uses real-time data and quantitative analysis (or math, as we say when we are not being fancy) to make better assessments of the value of a stock. And in the marketing industry, we can look at vast amounts of real-time data to judge the value of an impression. But generally speaking, marketers don’t.
In the decades of the digital revolution, the financial industry has doubled in size as a percentage of GDP. What about marketing? It is flat. Marketing is the same percentage of GDP as it was before the digital “revolution”. Marketing missed out. What did finance do differently? Managers in the finance industry hired as many mathematicians – or “quants”, short for quantitative analysts – as they could.
Price decided to see if she could outperform a typical advertising team by hiring some quants. After four weeks, Price’s team delivered ROIs from 50% to 100% greater than the conventional teams they competed against. And Price points out science is a process of continuous observation, revision, and improvement, so given time they could possibly do even better.
Process analysis showed a contrast between her team and the others. Marketing campaigns took a month to set up and only one in five days were spent on activities that improve performance. Marketers are saying that algorithms can do everything for you, but they actually still need quite a bit of human input. Like a robot vacuum stuck against the wall, algorithms require monitoring and occasional adjustment.
Furthermore, everyone in marketing talks about how great data is, but they do not seem to be aware of the dangers of noise. It is easy to draw the wrong conclusion from data if you don’t have great math skills. In one case, Price found that a campaign was “optimized” around a metric (clicks) that was not at all correlated with sales. In other words, it was not optimized at all;- the decisions – seemingly based on data – were actually totally random in their effect.
Price acknowledges that it is difficult and expensive to hire and train quantitative analysts. But there is a huge upside for the half-a-trillion dollar marketing industry. Increasing ROI by 50% could provide 250 billion dollars in additional value. And consumers would not be annoyed by an onslaught of irrelevant interruptive advertising. Using the data the consumers agree to share, marketers can stop bothering people with advertising and start delivering ads that feel more like a personalized service that delivers curated content that truly interests them. To get there, we just need to answer one question. As the kids say today – “do you even: math?”
