In the past couple of months, Google, Apple, and Facebook have all announced major changes to user acquisition tools, as well as the app store business and user experience.
These upgrades for smart acquisition bidding, discovery, and subscription have the potential to change the mobile game landscape for years to come. Matt Marshall, the founder of VentureBeat, delved into this at our GamesBeat 2016 conference a couple of weeks ago.
In a fireside chat, Marshall interviewed Lee Jones, the global product lead for mobile app advertising at Google, to discuss the possibilities that these developments open up for activating game communities and acquisition campaign design.
Here’s an edited transcript of their conversation. You can also watch the video embedded below.
GamesBeat: You started as a high school teacher, and then you moved up the ranks at Google. How did you do that?
Lee Jones: I started as a fifth grade history and English teacher. After that I went into a number of agency and startup jobs. A lot of them were in sales. People always asked me, in the interviews, “How did you go from being a teacher to being a salesperson?” I always said, “Selling history to 10-year-olds is harder than selling any of this crap.”
That’s the background. But I love teaching. My dad’s a teacher. That’s where I come from.
GamesBeat: Two billion downloads driven by Google ads. It’s an interesting number. This is a number you announced just a couple of months ago at Google IO. I want to set the stage and have you talk about the landscape in ads right now. Two billion is interesting because Facebook announced something similar, I think 2 billion plus. There’s definitely competition going on between you and Facebook.
Jones: We’re 2 billion plus as well.
GamesBeat: So you’re roughly equal. Which is interesting, because for a while most of us believed Facebook was the only shop in town when it came to driving downloads. Year in and year out, there were problems integrating with Google. Can you tell us how you did it? Where are those downloads coming from?
Jones: Just for quick background, I joined AdMob, which was the first company to put ads on the iPhone, back in 2008. I joined Google via AdMob. Back then AdMob was the leader in driving installs. We spent a lot of time trying to justify why you should pay for installs. It was so early. Gaming really led the charge. That year, Angry Birds—we used to say a Super Bowl happens every day on Angry Birds. I was an Angry Bird for Halloween that year.
GamesBeat: You had a story about Angry Birds, right?
Jones: My mother in law is one of the top 10 ranked Angry Birds players in the world. She’s not in your typical demo for games. It’s not the only game she plays. But she’s extremely good.
GamesBeat: Let’s turn her into a political statement for Google. She’s not in your typical demographic. My understanding is that Facebook’s main methodology is demographics, audiences, segmentation. You guys have decided you’re going after something else, which is the event. Can you talk about that?
Jones: Our entire thesis, what’s driving our product road map for UA through ads, is this assumption that there are so many different signals you can use to drive a high quality user. Every game is different. Your user base is not one specific archetype. And so the next big avenue, we believe, in paid app promotion is being able to share that unique insight and data with us about your users. We can go out and find more people that are likely to be valuable, like your most valuable customers.
We had a game, Pixonic, that we worked very closely with just on the AdMob side. They said, “Our most valuable customers are people who reach level five in our game. That’s all we care about. That’s the inflection point for someone who’ll end up being a paid user. Let’s optimize for installs from people who’ll reach level five.” We trained our system to do this and it resulted in a quadrupling of average revenue per user, as well as a 40 percent lift in downloads.
We believe that the signals and the insights that you know about your users are more valuable than any demographic. If all you’re doing is targeting men 18-24 or women 35-45, you miss out on someone like my mother in law, who probably spends a couple hundred bucks a year on IAP. That’s the primary focus for our ad products.
GamesBeat: I’ll bring up an inconvenient truth. Facebook just announced event-based advertising. Very recently, it was just last week or so. Could that be in response to the 2 billion number?
Jones: I don’t know. I can’t speak for them. But we’ve been doing this for more than a year. That product, the Pixonic example I gave, was based on optimizing for in-app event history and signals. We’ve had that product live in the marketplace for more than a year. At I/O we said, “Look, this has been going so well. The gaming industry has been at the forefront for helping us break through and improve from a user acquisition standpoint. Let’s expand that product to search YouTube and all the other assets at Google, and not just do it on AdMob.”
That was our big announcement at I/O. We’re currently in a beta with 200 customers.
GamesBeat: Let’s go back to the original question. How did you do it?
Jones: It’s been interesting to watch Google approach apps from their standpoint. Primarily we started as a search company. Since 2010 we’ve heard consistently from customers that apps are very complicated. Executing a campaign with search, display, or YouTube takes a lot of time. How can we make it easier? Really, what everyone cares most about is value.
If you flash forward to where we are today, we’ve put ads on search. We’ve put ads in the Android Play store. We’ve put ads on YouTube. We’ve expanded our AdMob publisher network to more than a million great apps. This is on iOS and Android. Obviously we don’t have the story on iOS.
Rather than trying to make you an expert on how to reach people across each one of our different properties, we also created this product called Universal App Campaigns. You can create one campaign to drive whatever your target goal is – whether it’s someone who reaches level 5 in a game, or if you just want volume of installs – and we will syndicate your ads, generate keywords, build creative, and run your ads for you.
We’re seeing fantastic results from this. One customer, I think it was HBO, said, “Previously it took us two hours to execute campaigns with you. We’ve got that down to 15 minutes.” They’re optimizing for subscriptions.
GamesBeat: Could you talk a bit about the feedback you’ve gotten from customers about the Universal App Campaign? If you’re going from two hours to 15 minutes, there’s a certain amount of trust. If the publisher isn’t doing the work and getting granular in those decisions, they’re putting a lot of trust on you to make those decisions for them. That could result in some questions about the ROI.
Jones: At first we thought Universal App Campaign, when we launched it, would primarily be for a smaller customer, a developer. We’ve actually seen adoption with some of our largest customers as well. They’re focused on the downstream event. Many customers will come to us and say, “Look, if you can hit my KPIs, we’ll spend as much as possible.”
Any marketer cares about where their ads show up, obviously. We still have a number of brand safety controls in effect. We show you reporting by channel. If there are keywords that are highly sensitive, you can exclude those. It’s the balance that we’re trying to find, but ultimately we think that the tradeoff between the time and the value that you get back is worth it.
GamesBeat: You mentioned that the new version of this Universal App Campaign is coming out soon. It’s in beta. What’s taking the time? There are a lot of bells and whistles you’ve been talking about – machine learning, these infinite numbers of signs you’re talking about, and the learning that feeds back into the product.
Jones: From my perspective it’s not really taking a long time. Typically, at Google, we’ll make a lot of announcements at I/O. We’ll bundle a lot of announcements and then have a commitment to launch those products within a certain time frame. So from my standpoint, this should be launching imminently. There are 200 customers in the beta. These are all of our top customers. If folks are interested in learning more we’re happy to consider other customers as well. This product is not officially available to all, but it’s pretty much available to our biggest customers.
GamesBeat: Can you tell us about the machine learning that goes into this, and some of the other artificial intelligence notions? These terms get thrown around a lot in the Valley. There’s some mystique behind it. But these things have been around for 20 years. Can you talk about the specifics of what this means? What are you doing that others aren’t?
Jones: First of all, we’re using Pied Piper for high video compression speeds. That’s been great. I highly recommend them. The whole premise here is, to reach a user at the right moment, there’s a lot you have to understand. First and foremost, it’s incumbent on the marketer, the gamer, the developer, to understand who your most valuable users are. Once you have that understanding, how do you get more of those people?
I saw a stat across all apps. On average, 20 percent of your users are driving 80 percent of your value. What if you could grow that 20 percent? We’re saying that one of the best ways to do that is to let Google leverage some of the great data we have, combined with your signals. It could be the amount of time people spend watching videos on YouTube. It could be have active gamer profiles we have from Android. It could be IAP spend. All of those signals we use to make a decision in real time about who to serve an ad to, what’s the cost, how we should bid, what keywords we should look at to maximize value for the customer.
We also generate creative. There are some things here, to your point, about machine learning and some of the big data clichés. We have a product called Trial Run Ads. I’m not sure if folks have heard of that. Because of our integration with the Play store, we can take your APK, your Android app, and actually build an ad that lets somebody try the app directly from the ad.
GamesBeat: Can you tell us how well that’s doing?
Jones: It’s also organic search results, too. You don’t have to pay for that. This can be enabled through organic search results on Google. As to how we’re doing, we haven’t released any stats yet. We’re expanding the beta, though, to include many more customers, which is exciting. Our plan is to put it in Universal App Campaign version two at some point.
Anecdotally, people are very happy with the results. Our first and foremost concern: is it a good user experience? All signs point to yes, if it’s served to the right audience. People prefer to try before they buy. Downloading an app is potentially an investment. It’s a data cost, if it’s a big app.
GamesBeat: I understand you can’t divulge too many details on the results of these things, but ads in the store, the Play store, are also relatively recent. There’s not a lot of publicity or chatter about that. Can you talk about the contribution that’s made? Apple has announced something coming down too for their App Store.
Jones: If you’re not aware, we put ads in the Android Play store. This was very exciting for us. Part of the impetus is that we want to make paid user acquisition affordable to all developers, not just the big guys who can afford to spend. That was a big driver here. How do we help make our Android ecosystem and user acquisition affordable?
As far as the numbers, the announcement of adding Universal App Campaigns and ads in the Play store resulted in a doubling, year over year, of the number of installs we’ve driven through ads. Tremendous impact. People see a lot of value here. For us, it was sort of like, “Hey, we’re a search engine company. We own the largest app store search engine. Why aren’t we putting ads there?” It was a no-brainer from a company standpoint. It’s also a decent user experience.
GamesBeat: Can you talk a bit about where you’re seeing some surprising upside in your traction? You have ads in the Play store. You have YouTube. There’s social ads and search. There are email campaigns. They’re a big driver. Can you talk about where you’re seeing surprising upside to driving installs?
Jones: From a gaming perspective, really good game marketing is when you’re able to extend that experience into your marketing campaigns. That’s been what we’ve tried to align with when it comes to building new ad formats and targeting emplacements.
YouTube is a great example. We have research that shows 90 percent of avid gamers visit YouTube once a week. They go there to watch other people play, to look for tips and tricks, and 64 percent of those people will download an app after seeing a YouTube ad. Just the fact that we’re able to tap into that community is amazing from a gaming perspective. The upside is tremendous. It’s actually a beneficial user experience as well.
I’d also say, from a Play standpoint, being able to be highly targeted as far as who you’re trying to reach reduces a lot of waste in marketing. We’re also able to do a fair amount of volume and drive quality users there.
GamesBeat: Messaging. Many people are saying that’s become the platform, with 2-3 billion people moving to WhatsApp, WeChat, Snapchat. A lot of people are playing games there. Can you talk about how you’re managing the swelling of people in these new communities?
Jones: There’s no doubt in my mind that messaging is big and will be big. What you’re seeing in the Asia Pacific region, this is already happening. Line and WeChat have vibrant gaming communities. You can do a lot just directly from the chat apps. The U.S. and Europe are a bit behind in that space. It’ll be interesting to see how this evolves. It’s still a bit early, I’d say.
From our standpoint, though, just overall in the industry, where we are as far as marketing—you no longer have to just try to drive the cheapest installs possible. If you can be smart and data-driven in your approach, you can unlock a ton of value. You don’t have to be a billion-dollar market cap. You can be two Neds in a shed and take advantage of this. That’s our goal.