The tantalizing promise of online advertising has always been to deliver the right message at the right time to the right person. But for all the big talk about "big data" and consumer targeting, finding the perfect message can still be a guessing game.
Tangerine executives knew this as they tried to figure out how to promote the bank's first ever money-back credit card, which launched this week. So rather than making one ad, they made 70.
The online campaign for the new card will be testing out dozens of different messages and images to attract potential consumers to its new rewards card. Some of them will focus on the specific cash-back percentage (2 per cent in two categories of purchases, and 1 per cent on all others), while others will communicate that same offer, not as a percentage but as "double" the usual rewards-card offerings. Other ads will focus more on the categories in which people can accrue rewards, depending on a person's interests – look up recipes, and you might be more likely to see an ad focusing on rewards for spending at grocery stores, for example; if you're a young person living in a city, you might be more likely to see rewards for restaurant spending. Tangerine will be using customer information from Environics to help target those ads, and use Google's dynamic targeting to test them.
The point of making so many different ads is that, as the campaign continues, the targeting can get better. If women in urban areas of Ontario respond better to a certain message, the bank will use customer data to deliver that same message to other women in those areas, or in other places which have similar characteristics. Messages that perform poorly could be jettisoned altogether.
"It's kind of like a puzzle," said Jennifer Matto, senior manager of media at Tangerine. "We're trying to find that perfect ad for everybody … but also get learnings for future advertising. For example, if we learn that home improvement is a really compelling category, we can take that learning into future commercials as well."
The concept behind this strategy isn't new – A/B testing has been around in advertising for decades, as smart marketers have experimented with different ad concepts running at the same time, in order to determine which works best. The difference here is the sheer scale of that testing. Because of programmatic advertising – automated ad buying that speeds up the process and can sometimes target people more specifically – and other advances in advertising technology, marketers are beginning to expand how they conceive of campaigns and test their effectiveness on the fly.
For consumers, it means that increasingly, if you are seeing a particular message, it's because a company knows something about what you like and who you are – or thinks that they do, based on your behaviour online.
Royal Bank of Canada recently took the same approach to advertise its Avion credit card, testing a number of different messages. Of all of them, the bank found that one – emphasizing that card members could book plane tickets without seat restrictions – was the highest performer. That ad had a 28 per cent higher rate of customer acquisition than an ad offering 15,000 bonus points to new customers, according to a recent report by Google.
The same report also noted how L'Oréal SA's Vichy brand saw results by using different ads to market its sunscreen products in France, depending on whether the ads were being seen by women with or without children.
Tangerine has already employed this strategy for ads about its savings products: recently it tested a few different images to use in its online ads and found that for savings, images of palm trees and sailboats got the biggest response. But running 70 ads, as Tangerine did, is a big leap from anything it has attempted before.
The tools to put art and words together in different combinations as ads are posted, depending on their audience, have only begun to be more widely used in recent years. Online ads have moved from static preset boxes to dynamic things that can be assembled in real time. As marketers become more comfortable with the technologies available to them, they are working more and more to make the targeting of their campaigns more precise.
"We're doing much more detailed optimization," said Brenda Rideout, Tangerine's chief strategy officer. "This is all done in real time. … We're in the continuous learning phase, but I think it's something we'll do with every campaign."