In-app native ads are seventeen times more clickable than banners. This startup wants to convert banner ads instantly to native
This is the story of how four college students from Cyprus developed an award-winning fashion app and then decided to pivot into mobile native advertising.
“We got some press, some users and we engaged them. They liked the app, but okay, how do we make money?” Christou reminisced to Geektime. Conno Christou, George Makkoulis, George Eracleous and Panos Papageorgiu are four friends from Cyprus who were studying together at Imperial College London in 2013, where they won an award for their newly created fashion app.
“Advertising was our only option. So we set to work creating six or seven technologies that could serve banner ads.”
Unsurprisingly, very few users clicked on these ads.
That’s why the team started Avocarrot in January 2013 with first investment from angels in the adtech space in London.
Avocarrot works with mobile app developers to generate native ads, which integrate seamlessly into the content of an app. Now Avocarrot, which has demo iOS and Android apps, has launched its latest product Morpheus, which takes banner ads and automatically converts them into native ads quickly and potentially at a massive scale.
“Native ads have a scalability issue because they have to be unique for every experience,” Christou told Geektime. “But with Morpheus there’s no need to create custom ad formats. We create native ads programmatically.”
Why native ads?
You may be familiar with the term native ads from the news industry. The idea is that a publication will publish an article that has the same look and feel as the rest of its content but label it as sponsored. In the mobile app world, native advertising means that the ad has the same look and feel as the surrounding content on the app. For example, suggested posts on Facebook are examples of native ads. So are promoted posts on Twitter.
Users tend to find this type of ad less annoying than banner ads, and may even share them, which almost never happens with banners. In fact, according to Avocarrot, native ads have 17 times the click through rate of banner ads.
What it takes to automate native ads
The only problem is that native ads are a lot of work. Every ad has to look a little different depending on the app it will appear in. Though Avocarrot’s algorithms customize these apps, for native advertising to be truly scalable, there needs to be yet another step of automation. Native ads need to be generated instantly and bid on in real time without any human intervention whatsoever.
“Morpheus transforms traditional banners into native on the fly as a commodity,” explains Christou.
Morpheus takes a banner ad, say 300 by 250 pixels, and adds more data that it takes from elsewhere in the Internet, say the number of downloads the app had or reviews. It also A/B tests more than one native ad format in each app, sticking with the one that converts users the best. The new native ads are served through AVX (Avocarrot Exchange) – the company’s own private ad exchange – which allows them to serve the highest-performing ads.
“We want to have a better click rate. It means more money for the publisher and it means that users actually liked the ad.”
Twenty percent of Avocarrot’s customers are brands, and 80 percent are apps. The company serves 200 million ads per month and works with ten DSPs or demand-side platforms — computer programs that purchase ads in an automated way. Each of these DSPs has hundreds of advertisers.
The company’s R&D offices are located in Athens, while the headquarters are in San Francisco. “In Greece you can get great programming talent at a fraction of the cost.”
A psychology experiment
According to recent reports, Millennials and people in the 18-24 age range are more willing to engage with native ads than their elders, even when they know it is sponsored content. At the same time, another study suggests that the more education a user has completed, the more resistant they are to sponsored content.
While both these studies refer to video and text that is embedded in a news website, presumably the high click rate of in-app native ads is based on the same psychology. Because the user likes and trusts the content they are engaging with, say on Facebook, they are more trusting of the ad integrated within it. A large part of that trust is due to the fact that the content is generated by actual people.
That is why Avocarrot and other native advertising companies are conducting an interesting psychology experiment. Can a computer mass produce ads that people will trust implicitly? Or will native ads start to get ignored once they “scale” and become ubiquitous? With payment per click so low, the ad industry needs a killer format like native ads, and the industry needs these ads to become scalable. But if that happens, there is a danger that automation will kill the secret sauce that made these ads successful in the first place.