Inmobi on Mobile Native ads: how to increase conversions?

Ashwin Venkatraman InMobi

Ashwin Venkatraman, the director of product management at InMobi gave an interview about native ads in AppInTop mobile app marketing podcast. Founded in 2007, InMobi is a global mobile advertising network, backed with $216 million in venture investment. Since launching its native ad network in April 2014 InMobi plans to invest $100 million over the next few years to develop its native ad platform.

All mobile traffic is concentrated on Facebook, Twitter, Instagram etc, the big social media services. Where does your traffic come from?

We have a network of mobile apps behind publishers; therefore we don’t actually depend on just our apps or properties for traffic. We work with tens of thousands of app developers and they all contribute to our traffic. The publishers we work with probably have hundreds of millions of downloads; they have worldwide distribution and really big apps. We consider ourselves a democratic platform. Everyone can work with us, use our products and get the best value out of it.

What distribution between your own traffic and the traffic from these SSPs do you have?

The majority of the traffic, roughly 90% of it, is from our own publishers, meaning that these are developers or mobile websites who work directly with InMobi. They carry our InMobi SDK within their apps.

There are partnership programs with traffic arbitrage and they offer a significant volume of installs in a particular country. Where do they get their traffic from?

A lot of these so-called traffic arbitrage or exchange or partnership programs tend to be happening in emerging markets. However, this is just a phase. It happened in the US and Western Europe as well, but the market matured quickly. People began to realise that when you’re in a barter system and there is no real currency at the backend, it penalises the person who is trying to get into this traffic. It’s still happening in markets such as the Chinese one where the market is not as mature as the US or any of the Western European ones.

Are there any anti-fraud algorithms that have been developed or are being developed that would work with mobile ad networks?

Absolutely. At a generic level, I would say every network needs to focus on making sure the network is clean and that they can maintain and build on the trust that advertisers have in them. At InMobi, we take this very seriously and we ensure that we have very sophisticated fraud detection algorithms for everything from validating ad requests that come to us to making sure the downstream conversions that are being reported or tracked are valid as well. These days, a lot of people spend a lot of time on their phones, but digital advertising revenue is not reflected in that. One reason for this is that advertisers are questioning whether it’s really that effective. Native ads are an answer to that, but we must make sure we give the right kind of user experience. Remove all the fraudulent activity and make sure the quality of every dollar you spend on native advertising is delivering the value that it should.

What do you think about ads based on the CPI model? Are they viable?

I’ll refer back to your earlier question about partnerships and arbitrage rate. A lot of people don’t want to take the risk of marketing spending and therefore optimise for ROI, so they go down the CPI path. This has two problems though. One, as an advertiser you are deciding that every install you get, that every potential user has the same exact value for you, which is most definitely not the case.

The second problem is that almost all of the risk of advertising and generating expected returns is being passed onto the ad network or publisher. This means that advertisers don’t have the scope or need to optimise their campaign. We don’t think CPI billing will give you the scale or quality that you can get by working with a lot of large networks. It’s probably why none of the top tier networks actually offer a 100% CPI-based billing system.

What is your average eCPM or CPC for different types of ad: native, video, and image ads?

This is going to widely vary based on the audience segment that you’re going after. For example, advertisers are willing to pay a lot more for the US audience. If you want to get a user in the US on an iOS device, given that the lifetime value you can generate from that user is much higher, they’re willing to pay much higher CPI, which translates to much higher eCPMs.

A lower LTV in a market that inherently has a lower pricing point, such as Thailand, will drive what the eCPMs and eCPCs are.

If you were to compare the ad formats though, we can see that native ads produce almost 5x higher CPMs for publishers, compared to regular banner ads. Video, which is another area InMobi is really invested in, generates almost 10  to 12 times higher eCPMs than banner ads do.

We are waiting to see what could happen if you combine video and native ads. It’s very exciting and we’re running some video lead tests on that.

What type of ads do you think developers should choose to maximise their revenue?

It’s very clear that the market is shifting away from banners. They’re intrusive and ugly. They’ve been like Post-It notes you just slap on top of your app. We’re telling developers; look, the entire premise of your app and this ecosystem is about design and clear experience. The latest phones are aesthetically pleasing with great user experiences at the OS and app level. Then you ruin it all with an ugly banner ad.

Native ads, video ads, interstitials are a much better way to show relevant advertising.

What is it about native ads that makes them worth investing millions of dollars into?

Every time a new medium of communication comes in, people generally start with what has worked for a different medium. When televisions were first invented, the first ads on there were just static looking newspaper ads with radio jingles. Suddenly, somebody realised, we should run video ads and it worked. Similarly when PCs came out, people just used print ads on them, then somebody realised that you could make the ads interactive which worked too. Mobile has evolved the same way. People put the PC banner ads on them, which don’t really work. This is where native ads come in.  We believe that building a mobile specific native format is what is really required to get mobile advertising to where it needs to be.

Mobile phones are very personal devices. It’s based on design and experience. The ad experience on them needs to be contextual, seamlessly integrated with what people are trying to consume and doesn’t get in the way of consumption of context like banner ads do.

You can get some conversations going because some people are frustrated with banners. At the end of the day, consumers will only make a decision if it makes sense to them. When we run native advertising units, for optimal campaigns we see about 3x to 4x better conversion rates and sometimes they’re as high as 5x better. They’re happening because people are actually interacting with the ads they find relevant.

In the video ad world, what do you think is the key to having a really good selling video ad in a native ad?

Mobile is the equivalent of a snack and video ads should reflect this. Make them short, catchy, bite-sized. The best performances we’ve seen from videos are those that are 15-20 seconds long. They clearly explain what the app is about and engage people. Show them examples of what your app is like, for example, if it’s a game, don’t show them the first level, show them more interesting levels. Show them all the features and what sort of experience they will get.

It’s also completely fine to experiment with it, fine-tune it. One of the great things you can do with data-driven advertising, that a lot of networks and InMobi offer, is retargeting. You can retarget the people who saw the first story or episode of your video. After a week, show them a different episode or show them more advanced levels of your game. Use a different kind of video messaging to reactivate people who have seen or used your app, but aren’t so engaged with it now.

What can App Publish do and where is your interest in it?

App Publish is a distribution platform that InMobi built based on one of our previous acquisitions. It makes Android app distribution to the hundreds of stores outside of Google Play, such as those in China, really easy. We’re integrated with about 150-160 app stores and you put your app on once, distribute it and it sits in all these app stores and you get distribution from them.

Let’s consider your traffic from a geographic standpoint. What is the breakdown by country and region?

We’re fairly global in the way our ad network traffic is distributed. It’s roughly, one-third Asia, one-third Europe and one-third in the North American market. We’re number one in China, one or two in Korea and India, among the top three or four in Japan. So in a lot of the emerging Asian markets, we’re very big. In North America, I think we’re in the top three and in Europe, we’re number one in a few markets like the UK and France.

There are some markets where we dominate, such as the Chinese market. We dominate there purely because we’ve been present there and have heavily invested in the market. We’re probably one of the very few global companies which has managed to crack it in the advertising space and go and build a business there. In a lot of the markets in Western Europe and the US, there is a lot of competition. We’re right there in the top, but there aren’t really big gaps between us and the next player.

What is your percentage of mistypes and how do you deal with that?

Within the general digital advertising rates, it’s hard to give you an exact number, but I would estimate anywhere between 10-20% would probably be unintentional or invalid clicks. You can catch them with creative innovation. One of the things we do is that we started putting in very explicit frames around our advertising when we do interstitials. Interstitial is a full-screen ad unit and we realised that if you put a full-screen ad unit, there’s not even like a really small space for your finger to touch the screen and not want to click the ad. Therefore, we made it 80% of the screen instead and put a clear frame around it. We started animating the way the ad enters the screen so that the user knows the game has stopped. The game is paused in the background and the ad is using 80% of the screen. The other 20% is transparent so they can see their paused game in the background. This way, you won’t keep clicking the ad hoping it will go away which is what people did with interstitials. They’re very frustrating. With our native looking interstitials, we’ve seen 30% better CTR, just because we made the experience better.

The way we started doing these with game developers was to say; let’s use characters from your game to present it. We have one of the largest word games, it’s like Scrabble and what happens is, we use the word tiles at the top of every ad we show. Therefore, as a user, you know exactly why we’re showing the ad and you know you’re still in the game because the tiles are from the game. It’s almost like you’ve done a very native product placement for an ad within the game.

The mobile app industry is growing and changing so fast.  Is your book, Apponomics, already outdated?

We’ve learnt a lot more since we published Appenomics and we will keep updating the content. InMobi runs a lot of developer initiatives.  Very recently we did a fairly large presentation at one of the PocketGamer conferences in Finland.  We continuously do webinars on our website. We write a lot of white papers, so we do a lot of things to share the knowledge that we get from every day of working in this amazing industry.  There’s a lot of content that we keep updated on our website.  We keep publishing new stuff so stay tuned for getting a lot, lot more in terms of what we can help our developers do for marketing their app.

Could you give me an example of a time where you walked into work, sat down and noticed something had already changed from yesterday?

Actually that happens a lot, sometimes more than once a day. The app ecosystem is largely governed by Apple and Android and they set the rules. An example of this is that Apple recently announced that you can’t run rewarded video ads where people get coins or in app currencies for watching them. For businesses that are only built on the back of serving rewarded video ads are suddenly in trouble. My advice is to not build  a business solely focusing on one niche because as Apple proved there, they can be shut down instantly. Fairly large players are fine, but the smaller ones can be out of business instantly just because Apple or Android changed something. This happens every few months.

It helps to keep your options open and work with people who can give you a choice so you’re taking less of a risk against the changing ecosystem.

What do you think the biggest challenge facing native ads is?

It’s very early in the development process. A few owned and operated properties such as Facebook have started using it. A lot of the people who started using native ads just did it on their own property. The real challenge is to democratise it and offer it to every developer.

The problem is not designing the ad. The problem is being able to convince every single advertiser to give you native looking assets that can fit into your native ad slot. Imagine, there are roughly a million apps on each store and each one of these guys comes up with their own format and goes and talks to the thousands of advertisers and says that each of them needs to come up with a lot of combinations of native ads. That won’t happen!

The advertising industry has standards and no one will keep creating custom looking ad units or creators for every single app developer as a publisher. This is where native ad platforms come in and we’re investing heavily to ensure that we give the tools and technology to allow any publisher to create customised and native ad experiences that are standardised enough to make sure you can access a lot of app developers, brands etc. All of these demands can actually run on your inventory.

The entire advertising space for networks is really a two-sided marketplace. There’s inherent inertia on both sides, which is where people like us come in who build the platform which bridges the two sides. While we give the customisation and design capabilities to our publishers to create native looking ads, we build tools on the other side which allow advertisers to actually create things. We help do the entire thing, which gets us 70% coverage in the market.

Looking forwards, what’s missing in the native ad world that you wish to improve? What is one feature you think needs to be included in the future of native ads?

We’re really excited about the kinds of creative possibilities that we’re seeing coming in to native ads. As of yet, nobody has figured out how to put video as a medium of communication inside native ads. Facebook has been trying it out on their app, but that’s it. This is a big and exciting challenge for us.

The second thing is to change the interaction on native ads. Mobile phones can be tilted, rotated, swiped, have a three finger tab etc. How can we design ad units that are interacted with completely differently because that’s what the background app actually does? This is another area we’re interested in.

Thirdly, which is another emerging feature of native ads, is how to make native ads seamless with the underlying concept of an app. A lot of content on mobiles is not sharable. They’re social by definition and how do we make ads social by definition? We want to make it so you can share the ad with your friend if you want to.

The industry is constantly changing and that’s really exciting.

This is an abridged and edited interview from the AppInTop mobile app marketing podcast produced by AppInTop, an automated mobile app marketing platform. Listen to this podcast episode or subscribe to the podcast via iTunes.