Native mobile ad landscape

My thoughts on it.

I’m tired of seeing banner and interstitial ads popping up every minute on mobile. They’re bunch of traditional web IAB units forced into tiny little screens, resulting in horrible user experience. Surely, there must be better ways to advertise on mobile apps. As I’ve dug deeper into the mobile ad landscape, I learned that 1) digital ad space is hella complicated, and 2) mobile ad landscape is shaping to be quite different than web, which has been largely powered by cookies.

When I was interning at Zynga in 2012, banners and interstitials accounted for something like 85% of all mobile ad impressions globally (I mean all of mobile, not just in Zynga games). That’s mind-boggling because they suck. Let’s start with banners. I can’t read their message most of the times because it’s tiny. The only time I click one is by accident because it overlays the app I’m using. Beyond the design factor, they are dumb. They serve random ads in a spray-and-pray fashion. No wonder CTR is negligible (avg < 0.4%). What’s mind boggling to me is how advertisers and mobile ad networks have worked around this. They launched mobile interstitials, increasing CTR at the expense of interrupting users from the app. Then longer versions of video interstitials came out to draw engagement with the ad — all in the name of increasing CTR and CPM without considering the users.

Money flows in despite the shortcomings of these ad units; mainly because no other forms are available in scale yet. It’s easy for developers to just take whatever is available to make money. Advertisers keep putting in the money because despite the negligible user engagement, cost is justified when you spray-and-pray in massive scale. It also goes on to show how difficult monetization is for mobile developers as many are willing to sacrifice user experience significantly for rev upside.

Below is a simplified version of today’s mobile ad ecosystem:

The obvious flow is the direct one between advertisers/agencies with publishers. Think of Coke working with its media agencies to advertise on Facebook. But publishers like Facebook are very limited. The rest of the long-tail mobile apps rely on ad networks (or ad exchanges) to serve ads. Ad networks connect publishers (apps that want to host ads) with demand (advertisers). In this long-tail ecosystem, advertisers never meet with publishers directly. Instead, advertisers provide the money and ad units, along with bidding criteria and target audience data through demand side platforms (they perform real-time bidding across multiple ad networks on behalf of advertisers). Through automated, real-time bidding and targeting, ads find their way to individual apps to get displayed. You can learn more about the basics behind mobile ad tech with simple Google query:mobile ad tech dsp ssp dmp.

Because there are so many publishers (apps), and advertisers, standard ad formats have been created to serve them in scale. Unfortunately, the standard came out to be traditional web ad units: banners and interstitials. Developers and users have no choice in selecting which ads get served in today’s model. Moving forward, I think three key changes will happen in mobile ad landscape

Publisher Advantage

On mobile, users spend the majority of their time on individual apps, not inside browsers. This creates the first problem on mobile; lack of scalable user identity/data. Without cookies tracking users as on the web, mobile needs to gain insights into audience targeting by other means. On web, providers of this data have been fragmented. Numerous platforms have their own cookie tracking to improve targeting, and data management platforms (DMPs) have emerged as extra layer for the sole purpose of adding/utilizing better targeting data. Even then, these platforms have no insights to the identity of the users. Everything has been around segmenting audience data to improve CTR/conversion odds.

So, when you have data/identity platforms like Facebook joining the mobile ad landscape, everything is about to change. FB ads work because they’re personal and more relevant to end users than competing ads. This makes sense because it builds ad products not only to serve advertisers, but also its end-users,taking user experience into the equation. There’s also just so much data about end users FB can leverage for better targeting and engagement. Compare this to other long-tail mobile apps that rely on traditional ad networks to serve random ads to their users. Facebook’s recent entrance to the mobile ad landscape via Facebook Audience Network means that the same powerful features can be brought to all other mobile apps now. I believe it also means that third party ad platforms like DMPs will become irrelevant. In fact, I think most of the intermediary mobile ad platforms will diminish (either dying or being bought up by large publishers).

I think this puts large publishers at an unbeatable advantage in becoming the data layer that powers ads across mobile. Facebook, Google, and Amazon are all key players with the potential of becoming market leaders. Each has already launched its own mobile ad network with native ad solutions. The good news for the rest of the publishers (apps) is that native mobile ad solutions that worked on existing publishers like Facebook will now be powering their own apps. The massive audience data can help these publishers with their user engagement and experience.

Amazon is another key example. The e-commerce giant uses their online-purchasing data to power its own product recommendations. Their mobile ad network will become impossible for advertisers to ignore as they begin to incorporate purchase behaviour data into mobile-ad targeting.

Big publishers like Facebook and Amazon will continue to make piles of money through direct monetization on their apps and via mature mobile ad networks, sharing more revenue with publishers (apps) hosting ads than ever before.

Serve Users

Second problem in mobile ads today is the inherent problem of ad tech focusing on serving the advertisers. It makes sense. Money comes from advertisers, and they are all looking to spend where ROI is the highest (measured by conversion, CTR, and CPM rates). Ad tech has favoured the buying platforms, increasing these metrics, and building around ad units advertisers are already familiar with (banners / interstitials) — all the while ignoring user experience. But with the push from large publishers building native ad solutions to serve their own users, mobile ad standard is changing.

As large publishers build ad solutions that are good for their end users, these solutions will start to emerge as new mobile ad standards. Native ads are great examples today. We’ve seen early solutions in publishers, like Twitter, scaling in mobile ad networks next (MoPub‘s native ad shown in pic below). The key here is to have mobile ad units that are good for the user experience but that can also scale to become market standard. I’m interested to find out what new sort of ad units will standardize on mobile next. Mobile-only ad units like sponsored push notifications, or try before you buy app reviews could be potential units. Whatever the case, I suspect the push for change will come from large publishers.

Now add all of the contextual data mobile unlocks; advertisers can find out the best moment to deliver the perfect message to its intended audience in a way that can delight them. Mobile ads may never provide stellar user experience, but they are going to suck much less.

Reengagement + Search

In the next few years, I think big opportunities in mobile ads will come from reengagement and search.

Typical smartphone’s home screen is occupied by more than 30 apps today, and each addition migrates user’s attention away. Most apps utilize push notifications to re-engage user, but push overload is going to drive users to opt out more and more. This sounds bad, but it just means opportunity on mobile ad will shift from customer acquisition to re-engagement. Deep-linking tech is a great enabler of this today, and I suspect new tech will emerge.

Deep-linking directs users to specific location inside a selected app. For instance, a search for restaurant will offer a link directly to that restaurant page in the OpenTable app if it’s installed, allowing users to make reservations immediately. Taking users directly to a particular content inside the app will improve conversion rates significantly. Because apps store identity and payment credentials already, utilizing deep-link ads enable 1-click payments (or any other user action), and reduce churn significantly. With the explosion of apps, developers are faced with the challenge of getting installed users to return and remain active within their apps. This is a huge opportunity for mobile ads.

Search is another big opportunity on mobile, although I think Apple and Google have the advantage of owning this market with iOS and Android. Consider one implication of search; today’s app store. How we discover games and apps is broken. More than 300M people visit app store every week only to discover the same handful of apps. What worked for 500 apps in 2008 is not proving to work for 1M+ apps in 2014.

Soon enough, mobile app ecosystem will experience the same shift from discovery of applications to intent-driven search, similar shift from Yahoo’s directory of links to Google’s search bar. But app store is just one implication of search. As contextual data like location gets incorporated into mobile, algorithms will understand user’s intent better and better; directing users to precise locations inside apps they are looking for. Along the way, I think fortunes will be made on ads utilizing user’s intent-driven search.