All Campaigns Are Not Created Equal
About 15 years ago, I was working for a little marketing agency creating banners and ads for websites. On average, a client had around 10-12 such banners. The rule of thumb was that if a banner was relevant to about 80% of the audience, we’d build it. A “20% off everything“ for everyone campaign would surely meet this criteria. Building a banner targeted to a handful of shoppers was simply not done, it was unthinkable.
The Era of A/B Testing
Around 10 years ago, the era of A/B testing began. Solutions appeared that were able to test various creatives against one another and determine which one performs best. These were exciting times. Retailers found out that a vast majority of their customers preferred yellow versus blue buttons and banners featuring a couple have a higher CTR on weekends than on Wednesdays. The number of creatives that made it to the website increased slightly and we saw the use of rotating banners rise to allow for more content within the same placement.
Although this approach brought more insight to how banners were created, it did not change the fundamentals. Retailers still chose those blanket “20% off campaigns“ but they did make the creative shift to yellow and not blue buttons and incorporated a smiling couple on weekend days.
The flaw of A/B testing is this: if you test two banners and find out that 68% prefer banner A versus banner B, and hence choose banner A, you will disappoint 32% of your audience.
The Personalization Era
Personalization of content solves exactly the problem described above. In the case of the two banners, it will serve out banner A to 68% of the shoppers and banner B to the other 32% that prefer it. This changes everything!
Now it makes sense to build campaigns that appeal to only a handful of people, because you can be sure only they will see it. Personalization means a retailer can now start producing hundreds or thousands of banners, ads and other pieces of content and will always be deliver the right message to the right audience.
Personalization also solves the problem of segments. My neighbors are 79 and 83 years old. One might think they’d be interested in gardening or books, but they are the most athletic people I know. They go sailing every weekend, she’s the president of the local tennis club and he heads up the local alpine ski group. Personalization will recognize these preferences based on their behavior and what products they’re interested in and will serve the right content, irrespective of their ages.
So are segments out of date?
Personalization judges the appeal of a content to a certain shopper by its performance. If you have a hundred banners all displaying a different products and one of them giving out a $200 discount code, each and every personalization engine will determine the $200 discount code to have the best performance. However, giving away money is not what economics is about. It’s about achieving most with least capital. So if there is a segment where you want to gain more market share, by all means create it and play the $200 card.
But for the rest, let personalization determine which content is right. And for the retailers and marketing agencies out there: start creating more content. We’ll make sure you hit your target.