Liveblogging Inside Social Apps: Acquiring Users on iOS and Android
We’re at the San Francisco Design center, blogging Inside Network’s third annual Inside Social Apps conference.
Kim-Mai Cutler hosts a panel on mobile marketing distribution featuring Fiksu CEO Micah Adler, Flurry’s VP Marketing Peter Farago, Google’s manager of global mobile sales and product strategy George Meredith and Tapjoy’s CEO Mihir Shah. The following is a paraphrased transcript of the discussion.
Kim-Mai: Apple sent out a warning to developers this week on manipulating chart rankings. Peter, what do you think the impact on the marketplace and the rankings is going to be?
Peter: What we’ve seen over the last several months is that these bought downloads have added pressure. If I go back to early Q3, to get into the top 25, it was maybe 50k downloads. There’s a lot of things in Q4 that added pressure — the 4S, the holidays — the bots added more to that. In January, we were seeing 75 and 80k downloads to get into the top 35 and even larger on weekends. Some of that is definitely [new iPhones]. But there has been substantial volume from these bots.
Kim-Mai: Now that developers can’t rely on these services, how will it change the types of apps and games that top the charts?
Peter: We look at the market from a 3 to 5 year approach. If you’re focusing on the consumer first and making an awesome app [it will get to the top of the charts]. We focus on helping developers understand the audience. The rules of doing efficient marketing spends haven’t gone away just because we’ve had a bump in the road. It shouldn’t change anybody’s business strategy.
Mihir: I think we should applaud the move. Offshore bots where a computer is gaming downloads to make users think an app is good is fundamentally damaging to all of us. My point is that the entire industry needs to get away from this velocity discussion. That we need a direct response component. It needs to be more about well-though-through spend to get your game out, test it, optimize it, to get more and then monetize it… Rather than 2011 and 2010, which were dominated by discussions of velocity.
Peter: There was never this problem on Android. It was “what’s your uninstall rate? Do users open the app? None of these things seem to matter on the iOS app. I think too many people are focused on getting to the rankings. [The should be] focused on building franchises to create a fan base, like Zynga does on Facebook.
Micah: Apple has taken a lot of steps on cracking down on impostor apps and now this, which I completely agree with.
Kim-Mai: What about changes to the ranking algorithm?
Micah: We haven’t seen any changes, but we have seen unnaturally high apps disappear. We work with all the different sources for downloads and we were approached by these services many months ago and we have tried it — it was presented as an ad network — but it become obvious to us that… users weren’t even launching the app. So we stopped using it after that. But it’s been this well known secret I guess in the ecosystem.
Kim-Mai: They issued a warning, but when you run a traditional campaign or a burst campaign, it can have an affect on the rankings. So where is the line with manipulation?
Micah: I don’t think there is a line. There is a clear difference between manipulation and influence. If you get a good review, everyone will download your app.
Mihir: I think burst campaigns are in the dated concept of velocity. I think going for a variety of high quality channels with partners that do it the right way — the work with other developers to cross promote, that get good write ups in TechCrunch — that takes time. Burst campaigns in my opinion is a thing of the past. Now it’s all marketing day after day. You go through a Fiksu, you go through a TapJoy, you go through a Flurry. It’s no different than any other medium, but we were spoiled by velocity. But most have become much more sophisticated in the portfolio strategy of getting a great app in the hands of users the same way any other media company does it.
George: Over the last four or five months the number of advertisers both big and small coming to us and saying I want to be on the top 25 or top 50 is really small. Most people are coming to us saying they want to start small and build up.
Kim-Mai: The index cost of a loyal user has gone up dramatically because apple had a great quarter. Is that going to accelerate more or decelerate now that the holidays are over?
Micah: Every month we publish a cost per loyal user and the average cost per loyal user — one who launches the app at least three times. I haven’t seen the January numbers so I can’t speculate. The December numbers were impacted by the holiday season and the late launch of the 4s. Q1 is always soft but the relative softness will be less than previous years.
Kim-Mai: What do comparable costs look like on Android? Is it half the cost or a third?
Micah: There’s a few differences between iOS and Android. On Android we can acquire downloads more cheaply and they tend to be more likely to engage with an app once they’ve downloaded. Android users for whatever reason are easier to turn into loyal users. On the other hand, Android users are harder to monetize –they’re less likely to pull out their wallet. If you define cost per loyal user as at least three launches then the direct acquisition costs are cheaper on Android than iOS.
Peter: We measure what the relative value of a user would be by platform — Android users tend to be worth less. It’s about 24 cents per user by in-app purchase on Android. We looked at apps that have versions on iOS and Android. If you made a dollar on iOS, you made 24 cents on Android. Everyone gets apps and tablets under the tree — we saw up to 7 million activations over Christmas alone. The one thing I would add to the discussion is that a company is going to want velocity. They’re just going to be smarter about the quality of users in the top 25. If you’re in the top, you’re going to get more downloads. But what is quality of users in the top 25? They’re promiscuous users – they try a lot. We see people that get 7K downloads on iOS just by being in the top 15. The market is correcting itself to where everybody is thinking about quality of the app.
Kim-Mai: How do you devise a marketing plan by platform?
George: We ask what kind of metrics they’re tracking. So many launch without basic tracking. For the budget, at this point a 50/50 is fine. The latest version of Android has a lot more monetization features. There are a lot of new things that are going to make monetization better. I see no reason to not think about both equally.
Mihir: If you’re an app developer and you have a game to launch, you’re going to have access to a tremendous user base. But you need to test first. I don’t want you to spend a lot first and then find your LTV is lower than what you spent. You need to take it in a very measured responsible way if you’re a new developer and you don’t have 20 other apps where you can cross-promote, then get to the arbitrage point and spend. In our second title types – games that we’ve helped develop, which accounts for 50% of our inventory – we work very closely with developers on virality and social and how to use virtual currency.
Kim-Mai: On Android you can use incentivized installs, though?
Mihir: I hear you but I disagree with the question. The goal for the developer should be a ROI positive experience where they connect with the user and build up a franchise that will grow. You can get scale by proving you have a great app that doesn’t get uninstalled. We have a strong point of view there. Very similar to what Google announced today, we opened up the bidding in our marketplace — the point of that is to allow advertisers to self-select at the bidding level that makes sense for them. After we did that our network jumped 20 percent in conversions, number of users participating and number of apps participating.
Kim-Mai: How do the marketing channels on Android differ from the channels on iOS?
Micah: Its important to have a lot of breadth. Advertising on Android is a very different experience and there’s much more impact from incentivized networks. The tracking is different and some networks are more opaque.
Kim-Mai: Assuming measuring is taken care of, how do you use marketing?
Micah: The tracking aspect that George brought up….
George: I would be cautious on incentivized on Android because nobody knows what the ranking formula on is on Android. Adding low quality users who may uninstall your app or never open it could actually hurt your ranking.
Kim-Mai: Why did you drop your minimum bid?
Mihir: We’re not an ad agency, we’re an open platform to connect users to advertisers. The issue of timing was simply one of scale. You get to an element of scale on the user side somewhere between 70 and 90 million MAU and that’s not impressions, that’s interactions with one of our ads. When you get to that scale, you create affinity groups. Not everyone there will want Dream Zoo. They may want communication apps. The big players can basically outbid everybody else. We want to introduce them based on our targeting in their affinity group. They have to self-select. Otherwise the first impression for everyone is going to be Zynga.
George: If search works or Google display works, it’s about performance. If your click is only worth three cents and if to somebody else things it’s four or five cents, it’s the maturity of the market. There’s enough volume in advertisers where that makes sense. Now that we’re getting closer to a unified system where the AdMob display ads are integrating with Google, it makes sense.
Micah: I applaud both of you for taking this step. It’s great for the marketplace, great for the ecosystem. This is a step toward a much more rational ecosystem. I think its good for publishers too because they can get those one cent clicks they wouldn’t have been able to get before.
Peter: We’ve been very open from day one. It’s an open marketplace so you can’t say to developers or companies “no you can’t do this, you can’t set the price you want.” I think at the end of the day, you have to measure the LVT of the user you acquired. If you look at a cost per click model your cost can be $7 or $8 dollars. I think the fact that people aren’t fully obsessed with ranking, being merchandised helps. We have a very long term view of the marketplace. If you tell users you can only work with me if you’re in my network, it impedes the growth of the market. We believe in targeting, we believe in finding the best consumer at the right price for the right app. You may find that you pay five cents only keep five out of 100 users. You might pay more. I think its great that Apple and the Android market are doing things that are better for the consumer. I think the problem comes when customers can’t choose. Constraints are things that ultimately impede the developer’s performance.
Kim-Mai: Facebook may make its mobile advertising debut soon. How will that change mobile display advertising?
George: I think the more people are selling mobile to advertisers, it opens up new markets to mobile. People are not just using the Facebook app, they’re spending more time with apps on TVs sometimes. It brings more advertisers into the space. We provide that network. I think its great. It’ll bring more advertisers to mobile. It’s the same with Apple. We saw better creative and it enhanced the game. We welcome it.
Kim-Mai: Obviously Facebook has strength in brand advertising, could they be a channel for getting users into apps?
Micah: For us, it’s great. It’s another partner to integrate with. Can’t wait ’til it happens.
Kim-Mai: It seems like the top charts are getting more volatile. Glu Mobile said the peak revenue period for games is shortening. What’s happened to people when they get to the top? What sorts of revenues are they seeing now compared to six months ago?
Micah: On iOS specifically it’s hard to maintain the top 25 position, harder than it was even six months ago without continued spend where you advertise across the network. If you stop, you will drop. That was not the case six months ago where you could get organic pick up. Android is very different.
Peter: There’s more competition for the top 25 spots. You can’t have the calcification of certain apps there forever because there’s only 25 top 25 spots and there’s a lot of competition for them. The industry talks about discovery a lot and solving that problem. Right now, ranked based discovery is the only thing on iOS. That’s where you have to find the right users. You look for recommendation, like on Android. There’s a lot of users that are qualified users that may not shop in the top 25 that you want to reach. The ROI is just not there. Now you’ve got to be disciplined.
Mihir: I think the fundamental difference we’re seeing as compared to AppData where people stay in the top 25 is we’re not seeing players dominating those positions. A year from now, you’re going to see people getting to scale in mobile with the ability to cross promote between titles and realize the social investments they’re making. They’re the guys that are the best in the biz. Increasingly you’ll see folks like Kabam where they share their games across their networks. GREE and DeNA do a great job of that. We’re in transition. I think then you’ll see the same publishers – not necessarily the game games – but the same companies in the top 25.
Kim-Mai: What are your thoughts on exits?
Mihir: We’ve got a lot of growing to do. We’re going to continue to build a great leadership company and then we’ll do great. Let’s try to capture value commercially.
Micah: I would expect there to be consolidation in the next year or two. There’s a lot of people coming into mobile with deep pockets. There’s a tremendous opportunity. We’ve tapped maybe one percent of the the market.
George: We’re in maybe the first inning. Mobile will become so much more. I don’t think there is a lot of pressure to sell. The opportunity has only begun to be realized.
Micah: Mobile advertising is one percent of what’s being spent on TV but in terms of engagement, it’s about the same on mobile as it is on TV if not more so.
Peter: Exit conversations are always funny because you’ve only go three ways to go and the only two good ways you go is to build value. If you look at smartphone adoption and compare it to PC and internet adoption — it’s been four times as fast than PC and twice as fast than the internet. There are probably 400 million active smartphones out there and if you look at the middle class in every country that can afford smartphones, it’s about 2 billion people. We’re less than a quarter of the way there for max install base and we haven’t had Madison Avenue dollars coming in. If there’s competition it’s artificially generated because there’s so much room to grow.