Advertising in the Data Age

In a previous life, I worked in online marketing. It is there that I learned about the value of data and the advertising industry’s insatiable appetite for it. Data is – and always has been – the currency of digital advertising, but until now, you weren’t the one getting paid.

Let’s rewind to the year 1994 when AT&T famously ran one of the first banner ads ever on the then freshly launched website of Wired Magazine:

One of the first banner ads (AT&T’s, 1994)

In 1994, this was a revolution. You didn’t need targeting or focus groups to launch a successful ad campaign, you just needed to be there. Today, the very same banner ad would drown in an ocean of digital noise, likely to go unnoticed by the consumer crowds who have long been numb to most forms of advertising.

The advertising industry’s initial response to falling click-through-rates was more aggressive ads. Popups, full-page interstitials and Flash soon made surfing the Web an experience not dissimilar to watching daytime television. The ever escalating struggle for consumer attention finally came to a preliminary end with the proliferation of ad-blockers.

Saved by data

This could very well have meant the collapse of digital advertising, but data came to the rescue. Rather than showing you more or flashier ads, advertisers soon came to realize the value of context, intent and timing. On the most trivial level, you were soon more likely to see an ad for “baby diapers” on a website dedicated to parenting than for example on a page about power tools. Consumers first became suspicious when ad platforms began tracking their behaviour across different websites. Soon you would see an ad for health insurance on a page about urban gardening, because you had previously done a web search for “why does my back hurt?” In 2013, I set my Facebook relationship status to “engaged” and soon after, I would see ads for wedding rings and other wedding paraphernalia all over the web. A year later, there was a suspicious rise in ads for mortgages.

The average click-through-rate (that is the number of people who click on an ad vs. those who have seen it without clicking) of a display ad is 0.06% [1]. That means it is 10 times more likely to be born with eyes that are two different colors [2] than to click on a banner ad. If, by using highly personal behavioural data, advertisers can bump that number by just a fraction of a percent, they more than double their revenue potential.

A losing game

The thing that worries most consumers is the non-transparency of it all. What data is being collected and who is it being shared with? Why do ads more and more seem to anticipate important life events, sometimes even before they happen?

“The billions of dollars spent to show you the most relevant ads will in the end only feed your advertising fatigue”

These growing concerns emerge at a time when things have already escalated in the advertising world. Ad platforms are amassing unbelievable amounts of personal data in an arms race to outsmart the competition. But this battle is being fought at the expense of the consumer and there is no doubt that it cannot be won. The billions of dollars spent to show you the most relevant ads will in the end only feed your advertising fatigue and make you suspicious of those trying to tout for your attention.

Another way

When we set out to build wysker, we started from the observation that “window shopping” really had no equivalent in the digital world. This inspired the unique wysker user experience that relies on a single button and presents products in a stunning full screen view. Early testers enjoyed browsing product stream after product stream at a rapid pace, but they also exhibited some interesting behaviours. Whenever they liked a product, they would unconsciously move the wysker button downwards to decrease the speed or they would let go of it completely. As we began collecting this behavioural data more systematically, we found that we could use it to paint a very clear picture of a user’s preferences and even their buying intent. After a 10 minute wysker session, we knew that a tester was interested in red sneakers, liked products that were somewhat sporty, had no interest in shoes from a certain brand and seemed to have a price range that ended at the $60 mark. We could determine all that because a) we had painstakingly tagged all of the products in our database with attributes describing their appearance and style and b) because in the app, we knew at all times what the user was looking at. The latter is fundamentally different from a web page where many things are displayed at once and what the user is actually looking at often is a complete mystery.

The second major discovery we made was that we didn’t have to make assumptions about a user’s intent. On the Web, you are just as likely to be shopping for new sneakers as you are to be conducting research for a paper about sea urchins. wysker is strictly a shopping app and so with every second you use it, you casually reveal what it is you are interested in buying.

All of this data is extremely precious because it is actionable. If you show an interest in red sneakers, the logical thing to do is to present you with a bunch of red sneakers to buy. This kind of data is precisely what advertisers are looking for. But it isn’t their data, nor is it ours. It is yours.

In the second part of this article, I’ll talk about how we keep your data private and how we leave it up to you to decide who to share it with.

Sources:
[1] https://www.thinkwithgoogle.com/intl/en-gb/planning-tool/display-benchmarks/
[2]
https://www.medicinenet.com/heterochromia_iridis/article.htm