Game companies, app developers, and mobile marketers are resigned to losing the Identifier for Advertisers (also known as IDFA), which has enabled advertisers to target users with relevant ads for years. Now those companies are figuring out what will happen next.
By getting rid of IDFA, Apple believes it can protect user privacy. But critics say it will dismantle the effectiveness of advertising, resulting in higher costs for users and layoffs at lots of game companies and mobile marketers. Those critics fear an apocalypse that could lead to a collapse of in-game and app revenues.
Starting in June, Apple announced that it would effectively get rid of the IDFA. It would do so in iOS 14, mainly by prominently asking whether users wanted to be tracked. The opt-in rates are estimated to be low, perhaps around 20%. Apple’s plan was to enhance privacy, but it caused apocalyptic fears among the likes of Facebook, mobile marketers, brands, and customers such as game developers. The fear is this change could stall a period of historic growth for games during the pandemic and force players to make purchases that they can’t afford right now. Apple did this without widespread consultation with the industry.
Developers complained. A few weeks ago, Apple relented just ahead of the release of iOS 14, saying it would not enforce the change until sometime in early 2021. So the mobile ad ecosystem has been scrambling to replace their traditional advertising tactics with something that could balance privacy and tracking. Mobile ad firm Vungle this week bought AlgoLift so that it could circumvent IDFA problems with “probabilistic” targeting in the first sign of consolidation related to the IDFA changes. It’s just one example, and probably not the last, of the change that the IDFA issue will bring to the mobile ad ecosystem.
I’ve been interviewing leaders in the ecosystem about what happened, why Apple went down this road, and what the solution could be. Our first part of the story explaining the situation is here. This is part two.
Clearly, some kind of compromise is necessary. It’s a tradeoff between effective performance advertising and user privacy. On privacy, the question is whether Apple can trust its third-party partners — and competitors — not to share user data inappropriately.
Eric Seufert, a user-acquisition and monetization expert and owner of Mobile Dev Memo, said in an interview that Apple had to delay the change because many popular apps would have stopped functioning, which would have resulted in a huge pushback from users and developers.
Fans of the retirement of the IDFA say it will protect user privacy. Critics say the changes will result in advertising spam, rather than targeted or relevant ads. Based on numbers shared in the Apple-Epic antitrust lawsuit, the developer share of iPhone revenues last year was $38.7 billion. If you take that down by 40% because mobile marketing is no longer effective, that’s a loss of $15.5 billion. Based on 2019 numbers, Apple’s own loss of revenue could be $6.6 billion. That’s a pretty big self-inflicted wound by Apple, and it’s why developers expect any replacement to mitigate the harm.
Sergio Serra, senior product manager at InMobi, said in an interview with GamesBeat that he wishes Apple just turned the IDFA into a rotating identifier, where it lasts for a few hours. That way, advertisers can find out quickly if an ad worked, but forget the user and the user’s data after that short period time.
“At least in that time frame, from a user acquisition perspective, I know whether my ad was effective,” Serra said. “This is what we mean when we say it makes sense for Apple to have an open conversation. Being the emblem of closed innovation, they will probably not listen. So we will need to reinvent the wheel. I don’t see advertising going away. But I see it taking a hit. And publishers will have a bias toward Android.”
We can say good riddance to ads. But the reality is that advertising is key to success for many games and apps, and if advertising disappears or becomes less effective, consumers will no longer get a lot of things for free. InMobi estimates that advertising is used as a revenue stream by 80% of app publishers.
Facebook, which accounts for the lion’s share of ad-related downloads on iOS according to AppsFlyer, said that developer revenue could fall by 50% without the IDFA targeting. Google and Unity Ads are also big players in the market. Julie Shumaker, a Unity vice president, said in an interview that its developer customers are certainly facing uncertainty on iOS and Unity is figuring out its response. Google has stayed quiet on the sidelines during the IDFA discussion.
Apple did not respond to multiple requests for comment on this story. In a public statement about the IDFA delay, Apple said, “We believe technology should protect users’ fundamental right to privacy, and that means giving users tools to understand which apps and websites may be sharing their data with other companies for advertising or advertising measurement purposes, as well as the tools to revoke permission for this tracking. When enabled, a system prompt will give users the ability to allow or reject that tracking on an app-by-app basis. We want to give developers the time they need to make the necessary changes, and as a result, the requirement to use this tracking permission will go into effect early next year.”
Now we’ll examine what happens next.
Feedback for Apple
Developers are presumably asking Apple for a stay of execution. They don’t expect that to work. But they’re disappointed that Apple hasn’t told them more about how to adapt. It hasn’t held a summit where it explains what will happen — even though mobile ads are a huge part of the iOS ecosystem and generate billions in revenues.
“My personal criticism is that Apple could have played this in a very different way,” said Offer Yehudai, the president of mobile app monetization firm Fyber, in an interview. “We know they don’t like to talk to others. But if you have such a big change, and even your top-grossing app publishers don’t know what’s going on, how do you manage this? Get us in a room, and spill out everything. Even if it comes out in January, we don’t know more than we did a month ago. It will still be buggy.”
Serra also said InMobi had no luck contacting Apple. Apple had planned to partially replace the functionality of the IDFA with an internal solution dubbed SKAd Network, a framework for measuring mobile ads with obfuscations that will protect user privacy. It takes away the function of measuring the effect of ads from third-party measurement firms such as Adjust, Singular, and AppsFlyer.
Some players in the ecosystem are testing new monetization strategies outside of ad targeting or retargeting. Adjust launched a new subscription app optimization service that will help advertisers diversify their portfolio and keep ahead of an ever-changing industry, said Adjust’s Katie Madding in an interview with GamesBeat.
Madding said that service subscribers may be more willing to share their identifier information than other users, based on the company’s initial research.
“We’ve been working with a ton of different clients at a global level to basically optimize something we’re calling a pre-permission prompt. We’re really focusing on our user experience expertise, giving our clients design tips in order to optimize toward that,” Madding said. “So we’ve seen for certain clients upwards of a 70% opt-in.”
That’s far better than the 20% opt-in many expect from users when Apple’s IDFA change takes effect. But it’s not clear if Apple is going to approve of these new methods. It has said that it will allow app companies to explain why they want users to opt-in. Apple is not, however, changing its own language about the opt-in question. And if the user says no to Apple’s question, it’s game over for getting that user’s data, Apple said in its clarification this week.
Apple’s SKAd Network
SKAd Network measures the results of advertising, but not in real time. It provides a report that doesn’t have user identifiers. When someone clicks on an ad and opens the store, the network will say which ad network, publisher, and campaign number led to the click. It will be sent after 24 hours and it will not have device or IDFA information. It severely limits the number of ad campaigns that can be run for each measured game or app. That is a big problem because advertisers can no longer scientifically measure the results of minute changes in advertisements, said Consumer Acquisition CEO Brian Bowman in an interview.
“You’re running tens of thousands of dollars or hundreds of thousands of dollars in ad spend, and you are flying blind,” Bowman said. “It’s not viable. The entire ecosystem we have evolved over 10 or 15 years is going to be rolled backward. In my view, the SKAd Network is a half-baked product.”
In a recent blog post, Bowman added, “Without the certainty of deterministic tracking or viable fingerprinting alternatives or time to process SKAdNetwork enhancements, the industry expects reductions in costs and ad spend associated with mobile app advertising.”
That is, the industry will pull back on advertising and possibly lay off user-acquisition staff in the face of uncertainty. Bowman believes that more creative advertising and more creative apps and games are the only real “levers” left that the industry can adopt in order to adapt to the changes.
SKAd Network also tells the advertiser 64 attributes of the user’s actions after the install, such as how far the user got on the first day.
Seufert said one thing that has to happen is that major networks such as Facebook have to now define the 64 attributes, so that the rest of the industry can use the same definitions and that a standard will emerge for how to do things with the SKAd Network.
The SKAd Network, once it is finished and bug-free, will move advertising away from user-level data — the idea of targeted, performance-based advertising to individuals to broader, market-based brand advertising.
This advertising plants a seed about a brand in your mind so that, when given the choice, you’ll buy that one over something else. It works, but it isn’t targeted. It’s like TV advertising, where the payoff happens, but it isn’t as clear. Apple’s move is generally good for brand advertisers, but not for indies who have no brands.
Instead of relying on mobile measurement firms, the industry will have to rely on Apple. And the interesting fact here is that Apple has its own advertising business in the form of Apple Search Ads. That business is growing, and it is likely to benefit, while Google and Facebook are likely to suffer. Apple has a mechanism where you can opt-out of sharing data for those Apple ads, but it is harder to find that then it is to find the IDFA opt-out setting. Currently, Facebook ad-related downloads on iOS could be 15 to 25 times bigger than Apple Search Ads, largely because user time spent on Facebook far outweighs user time spent in the App Store.
One of the benefits is that Apple will use cryptographic verifications that can’t be forged, and that should result in a decline of ad fraud. Serra of InMobi would like to believe that is true, but he doesn’t know for sure because Apple hasn’t described its approach.
Yehudai said that a group of companies banded together unofficially to understand what is happening, and the International Advertising Bureau agreed to standards for how games and apps to try to get everyone on the same page when it comes to ad standards.
Yehudai said that a set of privacy-friendly contextual parameters will help advertisers make their ads more effective. They can still measure things such as which country a user is in when they’re using an app, along with other details that don’t give someone’s personal privacy away.
“We came together with people we compete with day-to-day so that we can work this out,” Yehudai said.
Other companies are counting on “probabilistic” identification of users to figure out the likelihood that someone who is using a smartphone might be the one that clicked on an ad. This uses those contextual parameters that Yehudai talked about, but Serra at InMobi said that he doesn’t think Apple will allow that, based on a new FAQ it posted. Apple addressed a question, “Can I fingerprint or use signals from the device to try to identify the device or a user?” And Apple posted the answer, “No. Per the Developer Program License Agreement, you may not derive data from a device for the purpose of uniquely identifying it.”
In other words, some companies thought they found a loophole that would help make their targeting somewhat precise, base on probabilities and contextual information. But Apple has closed the loophole.
And the whole industry has yet to implement and train their AI models on these solutions.
“This is where they killed all hope. There was room for interpretation before, but now they’re clearly saying no,” Serra said. “This means they are listening to the ecosystem, but they’re still not actively helping or participating. If Apple is taking this strong stance, this means they intend to enforce it. That’s a huge problem for the publisher.”
If the ad industry pulls back and decides not to advertise, then the house could come crashing down, hurting lots of companies and leading to layoffs among ad and game companies alike, Yehudai said. But if everyone continues to advertise in the limited ways they can, based on the belief that some of the IDFA-free advertising will still work and gamers will continue to play, then it won’t the pain won’t be as severe, he said.
Winners and losers
“There are some winners and some losers out of this,” said Dan Barnes, chief operating officer at mobile game company N3twork, in an interview. “I think for the smaller developers, what they need to work out is that the best possible product is the one that has the lowest cost of user acquisition.”
In other words, developers will have to make a great game that is so good you don’t have to advertise it.
“As a developer right now, you should be thinking really clear about whether your art style, your messaging, your product pillars, your differentiation, is going to garner the lowest possible cost of acquisition for you,” Barnes said. “And then you should be thinking quite clearly about the first seven days of your product launch and how you are going to retain users.”
Samir El Agili, the president of New York-based game publisher Tilting Point, said in an interview with GamesBeat that games that require “micro-targeting” of users will be harder to grow without the IDFA. He also said user acquisition for mass-market games will be less effective.
“One task is to convince users they will have a better experience if they share their IDFA,” El Agili said. “We are looking at cross-promotion within our own games and looking at games in similar categories.”
With contextual parameters, El Agili said such advertising is not as efficient.
“The ability to develop other means to target users based on their behavior is something that we’ve done in the past. And we’re going to go back and do it a little bit more,” El Agili said. “I think the question for the whole industry is whether we will follow the same path, and how Facebook will adapt to this.”
Seufert thinks that the industry will see some casualties. He thinks hypercasual games — the ad-supported titles that take less than a minute to play and have become very popular in the past few years — are going to suffer as the value of the ads inside those games will drop to zero.
Smaller game companies are going to get hurt worse. Big game companies have lots of users already, and they can track their own users if they get permission from them, and so those big game companies can advertise their new games to people who play their older games. The big companies can also pick up new business by advertising third-party games to their users.
On top of that, the game companies that have players who spend tons of money in each game — dubbed whales in the industry — will also have a hard time because they can no longer precisely target those people, Seufert said.
“This hurts the people at the extreme ends, the people that run the games that are driven by whale monetization, and it hurts hypercasual,” Seufert said. “But those in the middle ground don’t get hurt as bad. They’re not doing the hyper precision targeting that requires all that user-level data.”
Whales are a huge source of revenue. If only 2% of users pay for something in a game, the weight of their spending can be so huge that it pays for the costs of supporting all the free players. But sometimes targeting whales is touchy. They may have spending or gambling addictions.
If ads are no longer successful in pushing apps and games to the top, we will see different kinds of apps and games leading the way. The winners will be the ones that organically rise because they’re so good and naturally viral, and Apple would probably like that, Seufert said. Gamers probably would like that too, but they may not like the fact they won’t get as much for free.
“That’s going to change the kind of games that are successful and the kind of games that get made,” Seufert said. “Apple doesn’t care about hypercasual games because it doesn’t make any money from them anyway. If I am a hypercasual game developer, this might kill me. I don’t think Apple cares for those people. If I’m a casual game developer and I have a broader user base, I don’t get hurt badly.”
Some developers might try to incentivize users with rewards if they will permit them to use their data. That’s OK to do on Android, but Apple hasn’t permitted that kind of thing in the past, and it is unlikely it would permit that in a privacy-first world, Barnes said. And, as mentioned, Apple has closed some loopholes in its rules on this.
“Apple definitely dictates terms here,” Seufert said. “So I don’t think that they’re going to walk any of this back. All the obfuscation is by design. That’s part of the privacy preservation initiative.”
And yet, in contrast to some of the others interviewed for this story, Seufert doesn’t think the failure to replace the IDFA with something good is going to cause a total meltdown.
“The obfuscation is the privacy protection,” he said. “It prevents people from doing any user-level tracking. There are limitations, but you can get around them. That’s my point. People exaggerate how bad this is going to be. A lot of [user acquisition] teams are overly dependent on the transparency that they think that the attribution companies provide. People complain about the limitations of the SKAd Network. It’s a lack of imagination. You have to be able to build the tools that make it performant. You can’t just rely upon third-party tools.”
In July, App Annie launched the market analytics product Ascend, which would make it easier for advertisers to understand the broader market and the relationships between products, such as whether strategy game players also like match-3 games.
Normally, mobile marketers help game and app developers execute user acquisition campaigns, where ad money is spent to acquire new users or to get them to spend money inside a game they already have. If a company can place an ad and see if a user takes an action, like downloading a game or buying an in-game item, then it knows how much it cost to get that user. Then it can track that user’s purchases and see how much it can spend on marketing to that user over time to see if it’s worth it. If it can’t track that particular user anymore, then it can’t make that calculation so easily anymore.
One thing advertisers will have to do is calculate how much they’re collectively spending on user acquisition campaigns and make sure that they get a payoff in collective revenue gains, said App Annie’s Ron Thomas in an interview with GamesBeat.
“Looking at these spending and revenue components together is a key tenet to be really successful in a post-IDFA world,” said Thomas, who’s App Annie’s general manager of analytics. “And we can’t just go deep in terms of every single user acquisition transaction on an individual level. We have to understand what the performance looks like on a macro level. You need to look at a campaign level and aggregate level of performance to see if you are allocating the right resources on the right types of campaigns and channels.”
Mobile marketing got over-indexed on individual performance, Thomas said. It’s not going to be an exact science anymore, and it will likely involve probabilities, he said. Those with powerful brands will benefit, and those with huge existing user bases will benefit because they already know who their customers are and they can precisely target them still.
One of the answers is to level up your creativity for your game and learn how to work with aggregated data, rather than user-level data.
“This isn’t the first time performance marketing has gone through a big change,” said Barnes. “We’ve gone through multiple of these over time. And ultimately, I think these moments normally lead to innovation. When you get out on the other side, it’s better for the consumer and it’s better for the advertiser. And I’m hopeful that that’s what will happen here.”
Brand advertising on a broad level can only get you so far. And that’s why Apple, Facebook, Google, the mobile marketing and advertising ecosystem, and game and app developers have to figure this out. Apple in particular has to recover some trust with game developers. And time is running out.
You can’t solo security
COVID-19 game security report: Learn the latest attack trends in gaming. Access here