The IDFA is dead. Is this the end of mobile marketing as we know it? At WWDC this year Apple essentially killed the IDFA — the identifier for advertisers. It’s not completely gone … but it’s now opt-in with a big scary warning.
In this edition of TechFirst with John Koetsier we chat with Eric Seufert. He’s was the VP of user acquisition for Rovio, he runs MobileDevMemo and QuantMar, and he’s a consultant. Now we’re wondering … will this kill modern mobile marketing as we know it? And Seufert says this is a “book of Revelations” stuff that’s going to reset virtually everything in mobile marketing.
Eric Seufert
What we talk about:
- Retargeting
- Mobile DSPS “are basically gone”
- Fingerprinting
- Eroding device graphs
- AEO (App Event Optimization) and VO (Value Optimization) on Facebook
- AEO and VO everywhere … on any ad network
- tROAS on Google
- Frontloading monetization
- Cohorts & segmenting
- Header bidding
- MMPs
- Deterministic and probabilistic attribution
- SKAdNetwork
- IDFV (Identifier for Vendors)
- Implications for app publishers and app consolidation
- And much more …
Listen: Death of the IDFA
Don’t forget to subscribe to TechFirst with John Koetsier wherever podcasts are published:
- Apple podcasts
- Google podcasts
- Spotify podcasts
- And multiple other platforms … see them all on Anchor
Watch: Death of the IDFA
Subscribe to my YouTube channel so you’ll get notified when I go live with future guests, or see the videos later.
Full transcript: Death of the IDFA
John Koetsier: The IDFA is dead. Is this the end of mobile marketing as we know it?
Welcome to TechFirst with John Koetsier. At Worldwide Developers Conference this year, Apple essentially killed the IDFA. It’s the identifier for advertisers, it’s not completely gone, but it’s now opt-in with a big, scary warning. And we’re wondering will this kill modern mobile marketing as we know it? If so, it’ll be kind of interesting because nobody knew about this thing, except for insiders in the industry, as of even a week ago or something like that.
To chat about that, I am going to bring in Eric Seufert. Eric Seufert has a long history in mobile, if you’re in mobile marketing you know his name. He was the VP of user acquisition for Rovio, he runs Mobile Dev Memo, he runs QuantMar.com, he’s a consultant, he’s a trainer, he does a lot of stuff. Welcome, Eric!
Eric Seufert: Hi, how are you?
John Koetsier: I’m doing well, thank you. You’re doing well as well. You’re in Texas, I’m in Vancouver, Canada. Let’s just get right into it. Eric, is this a mobile marketing apocalypse?
Eric Seufert: Yeah, it is, it is, this is book of Revelation stuff. This is like a new era, new paradigm, rebuilding the entire tech stack that you have been operating with and hoping that you can sort of apply that to a similar looking strategy, right? This is a big deal. It’s like an earthquake, I mean, it’s a massive shakeup.
John Koetsier: No question. And not just because Apple has done it, not just because Apple has the most valuable users in the mobile ecosystem so you can’t ignore it, but also because of what others might do as a result, where Google might go at some point and other things like that. But let’s go down a list. I mean, like what dies as a result of this? Do lookalike audiences die?
Eric Seufert: I believe so. So I think, do we kind of give like a high level overview of what has happened or we just assume everybody knows.
John Koetsier: Let’s do that. I’m probably assuming too much here because I wrote about it on Forbes. It’s gotten a lot of attention, over 300,000 views so far. Why don’t you tell people what the IDFA is and what Apple did.
Eric Seufert: Cool. So, back when I started in mobile in like 2010, 2011, Apple had a device identifier called the UUID, which was a unique device ID that was attached to every single device that apps had access to, and could use to basically store in-app behaviors against, right. So if I’m thinking about like, I’ve got a user in my app and I want to track what they do over time, then I’m going to use that ID as their unique identifier within my own analytics, track their sort of progress in the game for instance, track how many gold coins they had or whatever, but that would be the basic ID that I had, right?
And then when people started promoting their apps through paid user acquisition, these companies called attribution providers emerged, and they sort of allowed you to sort of attribute an app install to an ad campaign via the UUID. That was the connection that those were the touch points, right? That was a touch point, and so basically when you click on an ad you go to the App Store, all the data about what you had just done, your behavior is lost. There’s no, like the URL basically gets truncated, the App Store opens up then you download the app. To the app developer you’re just some totally random person, they had no way of identifying that person as the person who just clicked the ad.
And so this data, attribution providers figured out a clever way to sort of capture that UUID at the click, and then when the app got open, capture it again and then they’d say, hey, well we just saw you and that click was how you got into the app. And so we credit that click to that advertising campaign, and then that advertising network gets paid, and on and on.
John Koetsier: Exactly, and because that UUID was device-based and therefore permanent, that was a massive privacy concern, right?
Eric Seufert: Yeah.
John Koetsier: Which is why Apple created the IDFA, which was something that could be reset or even turned off.
Eric Seufert: Right, exactly. So they created the IDFA, which is basically like an advertising identifier that’s attached to the device’s unique identifier that’s attached to that device, but they gave users the opportunity to zero it out. So you could just say, hey, I don’t want to be tracked with ads, so my identifier is just all zeros.
And then you were basically anonymous because there’s no way to associate that with any given device and they also allow you to reset it. So okay I like being tracked with ads, but I want the ability at any given point in time to sort of reset this and erase anybody’s historical view of me as an ad viewer or whatever. And so that was more privacy centric, and so basically — and that was iOS 11, right — so that was a couple of years ago they introduced those features.
So here we are WWDC 2020 and I, so back in February, I mean, I think a lot of people had seen this coming, this change, right? So back in February, I wrote this post called “Apocalypse Soon,” and I walked through this hypothetical situation wherein Apple at WWDC20 says, hey, look, we’re making this limit ad tracking feature, we’re making it on by default.’
Right?
So it will be on by default and every user is LAT, like limit ad tracking by default and so therefore, essentially everybody’s ad ID is zeros and so everybody’s anonymous, and we basically remove the ability to do ad tracking. And then I kind of walked through what would happen after that in the post, and this was back in February. And so on Monday, Apple holds WWDC and they announce, hey, well, they didn’t announce that they’re putting LAT on by default, but they said, hey, from now on, starting with iOS 14 which launches in September, every single app has to ask a user if they want to opt in for limit ad tracking for that specific app, right?
And the example they gave is a pop-up and the wording was just like so extreme, it was like, so harrowing.
John Koetsier: Yes, would you like this app to be able to track you, know your banking information, social security number? Haha…
Eric Seufert: Right. So of course looking at that example, no one would click yes, opt-in, right?
So I think the idea here is that, my sense is that the opt-in rates on this are going to be just vanishingly low. I mean minuscule, like, oh trivially low, basically round down to zero. And every device therefore will not be trackable at the user level.
And in parallel to this, Apple last WWDC — or a couple of years ago actually, Apple announced, or they didn’t announce it, it was very quiet — they released their documentation for this thing called SKAdNetwork, which basically is a way for Apple to do attribution within the App Store. So Apple handles the attribution, they interface directly with the network, there’s no need to have an MMP like a middle layer that relays this data back to the advertiser, it just gets done via Apple. And so they, with this WWDC they added some new features to SKAdNetwork, and so they basically said, ‘Look, there’s no more user level ad tracking, but we’ve now enabled SKAdNetwork to track the campaign ID and the publisher ID from a click. So this click will get sent and then we’ll send all that info back to the ad network, we’ll interface directly with the ad networks and we won’t tell the ad network who you are.
You still won’t get user level data around ad clicks, or installs, or conversions in the app. But we’ll tell the ad network, A, that someone downloaded the game via that ad campaign from this publisher, which is valuable, and also if there’s a conversion that happens with 24 hours we’ll send that data along, right?
John Koetsier: Exactly.
Eric Seufert: So basically it’s anonymized, aggregated at the campaign level, but they still send some of this conversion data back. So that’s what happened.
John Koetsier: And the reason they did that, just so people have the context here as well, is that this is about an $80 billion industry in terms of global spend in 2020, estimated global spend in 2020, for people who are paying for user acquisition in mobile. So it’s a big deal and you’ve got to know what’s happening. If you’re going to spend money on an ad campaign, you need to know is it working or is it not working? So this is one thing that Apple has put out there, but this IDFA is basically going to be gone as soon as iOS 14 is out. You’re basically, like you said, is down to zero and that has implications obviously for mobile marketers at point of acquisition, but it also has implications in terms of retargeting. It has implications in terms of how ad networks can find new users for you based on what you’ve got in your app right now. I mean, what do you think dies as a result of this?
Eric Seufert: Right, so what dies? I think, retargeting is just gone, right? There’s no way to retarget a user without having a device ID, without having an advertising ID. I think it’s just done. There’s no way to do it, except unless you’re talking about you having like basically PII, you’re having personally identifying information about the user that maps to what Facebook has, right, and then you could sort of retarget them. So like, if I have users email, right? So if a user is in my game or my app and they’ve registered an account, and they give me their email, and they give me, they volunteer all this data to me, right? This isn’t exploitative, they created an account, they used their email address. I can retarget on Facebook with email, but that’s not nearly as broad…
John Koetsier: And actually that’s a question as well, honestly, because in some of the WWDC sessions, Apple has kind of skated around saying if somebody said, ‘No, don’t track me’ that is kind of global, right?
Eric Seufert: Sure, yeah.
John Koetsier: It’s not something they could stop, but it is something that they would probably look at and not be too happy about.
Eric Seufert: Right. Yeah. I mean, that goes against the spirit of the opt-in/opt-out, right? So hey don’t track me, but you know, I’m sending you a retargeting ad then, you know, hey, wait a minute, obviously they’re tracking me. So that could blow up in someone’s face, but you could do it that way. But also email, that’s not as reliable, right? Device IDs obviously they’re, you know, I know that this person has this device,it’s like this immutable property of them and so I can reach them on that device ID. Well email, I don’t know, like, are you going to be on your mobile phone when I retarget you? With this mobile gaming ad, for instance, you could be on your desktop, right? Or, you know what I mean. Or you’ve changed emails, like it’s an old Hotmail account from back in the day or whatever.
Because I mean, Facebook, I haven’t updated my email on Facebook in forever. I’ve had multiple emails since I first registered that account whenever it was, so it’s not as up to date, it’s not as reliable. But I think, so anyway, I think retargeting is mostly dead here. I think mobile DSPs are basically gone. Now I think there’s going to be some runway here for them because they have device graphs, right? So people, even if they’ve opted out…
John Koetsier: Eroding device graphs.
Eric Seufert: The what?
John Koetsier: Eroding device graphs.
Eric Seufert: Right, eroding, and they’ll erode over the course of the next however many months that it takes iOS 14 to get to scale. But also, I mean, they still have them and even after, so when iOS 14 is at scale, unless someone’s gone LAT after having their device be in this graph, they’re still kind of targetable on the DSP, but that’ll erode over time. That’s not like a long term value, you know, valuable asset. Yeah, I think so retargeting…
John Koetsier: What about fingerprinting?
Eric Seufert: Fingerprinting was never reliable. I think that’s like a myth. I don’t think fingerprinting was ever reliable. I don’t think it was ever like, you know, functional beyond… so there’s a lot of problems with fingerprinting. But I mean, it was first of all, just the fact that it was always a backup, right? I mean it was never meant to be the sort of first tool in doing attribution, but also I think people just vastly overstated how accurate fingerprinting is. I don’t think it’s accurate at all.
John Koetsier: Absolutely. And besides, you could very easily argue that not only is that against the spirit of the do-not-track stuff that Apple is doing, but also that Apple will probably actively take steps to make fingerprinting pretty useless. I mean, they’ve done that with ITP in Safari, intelligent tracking prevention in years past, and they’ve clearly stated that fingerprinting is against the spirit of the rules in some of the WWDC sessions in 2020. And they may make all iOS devices look very similar in some way, or give spurious input or something like that, that changes regularly and so you would not be able to accurately fingerprint.
Eric Seufert: Yeah, yeah. I don’t think fingerprinting is going to be viable so fingerprinting goes away. I think lookalike audiences, same thing, right? Like lookalike audiences, you could build a lookalike audience in the same way that you could retarget on Facebook if you have PII. But again, like it goes against the spirit of this whole thing, and also it’s just not as reliable. The device ID is just like the immutable one-to-one, unchanging sort of feature of any given user. Anything else they give you is going to be messy, right?
John Koetsier: Yup, yup. What do you think about AEO and VO on Facebook?
Eric Seufert: So I think, so Facebook it’s hard to say … just broadly, I think it’s going to be difficult to sort of evaluate here because you don’t know what kind of deal they’re going to be able to strike with Apple, right? Let’s say that Facebook is on similar footing to everybody else — which actually, I think they might be, because I was hearing rumors last week that Facebook was one of the first sort of beta partners for SKAdNetwork — so if they’re running FAN through SKAdNetwork, then that probably means they don’t have a privileged position.
So I wouldn’t be totally surprised if they don’t, if they’re just on equal footing with everybody else, but they maybe got first access to it or something to build some infrastructure around it. So AEO doesn’t go away, and actually I think this is the positive side effect of SKAdNetwork, now the other ad networks get to build out AEO campaign strategies because before they could have, right? Before they could receive these events from advertisers, but nobody sent these events to ad networks because they didn’t trust them. They didn’t want to give them, hey, here’s my really valuable user who just made a payment, who makes payments in my games regularly, here’s their ID. They didn’t want to send that data to ad networks because they didn’t want the ad networks using that data for their other clients, because they’re just going to turn around. They’ve got a device graph, they’re going to say, hey, here’s a paying user, I’m going to show your ads to him because he’s more likely to click and pay, and then I get more money from you as an advertiser. Right?
So I think now there’s no incentive for advertisers to withhold that data from ad networks now because it’s anonymized, so they can send this data back.
John Koetsier: Yes, indeed. The interesting thing there, and we’re talking by the way, we didn’t define our terms here, but basically AEO and VO are ways of finding users who look similar to users you like who are doing well on your app. Different ways at Facebook, but the data that you’re going to get back from SKAdNetwork, essentially from Apple, is only in a very, very limited time span. I mean, you’re talking 24 hours, that clock can be extended a little bit, but probably not indefinitely. And it’s only a couple parameters, it’s not a lot of data about who are the best users that you’ve got, what they’re doing, it’s a couple parameters there, correct?
Eric Seufert: Right. It’s a conversion parameter that you can set with a value between 0 and 99 that you have mapped to specific events in your app.
John Koetsier: Yeah.
Eric Seufert: And so, and also it can only be overwritten by values that are higher than the current value, right? So you think about, okay, if I’ve got value zero is a finished level one. The value of one is finished level five or whatever, and I just have some values assigned to these various different things. You’ve got, they have to be sort of in ascending order of value to you. But if let’s say that value ten is someone purchases an IAP for a dollar or whatever, IAP1, like the lowest value IAP. If they do that five times, you only get that value one time. So that’s what prevents a VO type campaign from running in this, because you don’t actually get that magnitude of spend. You’re only going to get an indicator that they did that spend once.
John Koetsier: And you’re going to have to be really smart in your own app about when you want to send that and what data you want to send that with. So you’re going to have, you’ve basically got one chance to send it, if I understand correctly. And you’re going to have to send it fairly soon, probably in the first 24 hours but it can be extended. But you want to say, you know, here’s a major conversion event or something like that that tells you more about that user. If we talk about, what about target ROAS, or return on ad spend for Google Ads? What do you think the impact is there?
Eric Seufert: Yeah, I mean, I don’t know what they’re going to do. They’re going to have to try to… so, what I think this does for advertisers is this is going to incentivize people to front load that monetization in the first 24 hours of app open, right?
John Koetsier: Yes.
Eric Seufert: You’re going to want to try to get as much there as possible to capture as much of the LTV in that first 24 hours, because then you can kind of get some sort of ROAS indicator from the events that you’re sending back. And so I think this is going to change product design in a major way. I think for Google, I will say Google the tROAS campaigns work on one or seven days anyway, so now you basically have one day, right? Same with VO though, and so you’re going to have to try to capture as much value in that first day. And then Google is just going to have to say, well, look, we don’t really know what the, we’re not going to get subsequent events, so we don’t really know, we’re not going to have any sort of underlying sense of how much these people are going to spend overall, but we’ll just try to get as much as we can in the first day.
John Koetsier: Yes.
Eric Seufert: But the other part of this is like, they don’t know… so all this stuff is aggregated, right? There’s no user level data. So when people say, okay, I’m experimenting with who to show ads to, because I want to make sure that — so they’re basically like in a day they may have shown, may have gotten a hundred conversions back and shown the ad to a thousand people or 10,000 or a million people, right — well, it’s kind of impossible to say who those people were that sent the conversion event back.
John Koetsier: Yes.
Eric Seufert: They don’t know. They’re just going to have to say like, well, okay, on this day, I targeted these types of people and I got more conversion events and then I sort of shifted my conversion, my targeting to this group of people and I got fewer. So group A was a better targeted group, but I have no idea who within this, I can’t dive deep into that.
John Koetsier: How you do cohorts, how you do segmenting, a lot of stuff will change there. I mean, you do know what ad campaign it was from, so you can do some cohorting or some segmenting on that, but the interesting thing that I thought about as well is what about ad based monetization? Valuing ads becomes more challenging, correct?
Eric Seufert: Yeah, and I think header bidding might be sort of impacted adversely from this as well, because how do you value a user, right? Like how do you value this any given user, which header bidding is meant to do, it’s a unified auction between all these different suppliers. How do you value that specific user without the device graph, without user IDs.
So I don’t know, I mean, header bidding might be going that way as well. I think it’s, and monetization, all this stuff, it’s like in the span of a day everyone’s models went from bottoms up — like, hey, let’s build device graphs, let’s try to evaluate users on their specific merits or get kind of granular with dimensionality — it went from that, to just like we’ve got to go tops down. Like we have, we’re a very, very broad categorization and segmentation of users and that’s how the models have to work now. And also that in the span of three months, because this rolls out in September.
John Koetsier: Yep, yeah. It’s a big, big deal. You’ve talked about MMPs and you’ve talked about the role of MMPs with SKAdNetwork. Obviously Apple is taking charge of attributions at this point, not giving you device level data or personally identifiable information, but it’s telling you that an attribution occurred, that a conversion occurred and it’s cryptographically signing that. So it’s telling you what’s going on there.
You’ve said that takes the MMP out of the picture. I’m not a hundred percent certain there, because as an advertiser, you’ve still got to trust that your ad network is telling the truth when they tell you that, hey, Apple told us that we have a thousand new installs for you, so we’re billing you for those today. And guess what, if you’re dealing with multiple ad networks, you know, how do you still get that information together?
Eric Seufert: Yeah, yeah. Well, I mean, there’s some level of trust anyway. I think, so here’s where I think MMPs don’t have a position at the center of this dynamic anymore. Now, but what MMPs are perfectly positioned to do is be the, sort of provide the probabilistic guidance around where money should be spent. Now that means because the ad network’s going to get the attribution directly anyway, right? So MMPs are just no longer in that sort of transmission circuit.
But if then that gets relayed back to the MMP either from the ad network, which I’m not sure how that would work, or from the advertiser, which I don’t know how the ad networks are going to… so, if the ad network gets the SKAdNetwork payload and do they immediately then transmit that to the advertiser and that gets ingested via API by the advertiser? I don’t know, I mean a lot of stuff has to be worked out, so if that’s the case, then does the advertiser then transmit that to the MMP? I mean, this is like a multi-node chain.
But let’s say that that happens, I think that MMPs are in a perfect position to collect all this data and tell you, hey, you get more incrementality. You get more incremental value out of your spend when you have 30% of your budget on network A, and 50% of your budget on network B, and on and on, split across however many networks than you do in some other configuration of spend distribution.
John Koetsier: Yes.
Eric Seufert: Because they have so much historical data. No one else has this historical data set, right? So I think the MMPs are perfectly positioned to be that kind of, to provide that guidance. And then another thing, another big opportunity here is, if you think about okay, what do you do if all this stuff is aggregated at the level of the campaign, what are advertisers going to do? Well, they’re going to make a million different campaigns, right?
Now SKAdNetwork only allows you to have a hundred different… so I think I misspoke before, I think I said SKAdNetwork allows you to have a hundred events, no it’s 64 events. They allow you to have 64 event tags and a hundred campaign ID tags. So what are you going to do? You’re going to max that out. You’re going to make campaigns that are very specific so that you can at least try to, if you’re only able to value things at the campaign level, you want those to be super specific, right?
So then I think what an MMP then is probably well set up to do is help you automate all that. You’re gonna have to automate…
John Koetsier: Exactly, understand where those are and take those numbers, translate them into human readable stuff, give you stats and data on how your campaigns went. That makes a ton of sense. Let’s talk about, and you’ve mentioned a little bit as well, obviously Facebook and Google are the giants, right? We’ve talked about Amazon becoming bigger. We know Apple Search Ads is significant on iOS and maybe this enhances their ability as well.
But Facebook and Google are the giants here. Does this really level the field somewhat in terms of amount of data that other ad networks can get, and maybe reduce the mobile user acquisition markets dependence on Facebook and Google?
Eric Seufert: It might, it might. I’m always reticent to say, hey, this is going to hurt Facebook’s position because they always seem to find a way to come out on top. It might, because they’re all working on the same dataset and their device graph and all the sort of demo data that Facebook has that’s attached to a device ID, sort of no longer relevant because it can’t be used as easily in targeting.
So it might do that. I think what it also might do is if someone says, look, I can only, I’ve got… a $100k to spend a month or something, you know, millions in a month, I’m going to spend $200k on… $500k on Facebook, $100k on five other networks. Well, it’s like, well, okay, if I’m evaluating the efficiency at the campaign level and I’ve got it spread across five different networks, and I know Facebook could take the entire million, maybe I just go with Facebook. Because Facebook is going to be doing a lot of that analysis for me, like within Ads Manager they’re going to be telling me like, hey, well you spent this much and you got this many events back, and so this was the more efficient campaign.
And so like, well, if I know that if I can’t track these, any of my users within my own system to the campaign ID to get LTV by campaign, I might as well just let the biggest network take the most money and then use our own internal sort of systems to optimize the spend distribution. So it might be such that when people are like, well look, you know, I could try to distribute this across a bunch of different networks, but I’m disincentivized from doing that because they don’t, you know, there’s economies of scale here. Why don’t I just give Facebook everything? Like it’s so much easier, and then I don’t need to use any third party tools. Now it’s just literally it’s just Facebook.
John Koetsier: And of course if everybody tries that, all of a sudden Facebook ad prices go way the heck up and Google becomes more interesting, and other partners. So the system balances and rebalances and rebalances. We have a question here from Justin Fibich, he says, “Eric Seufert, what happens to hyper-casual game publications or publishers?”
Eric Seufert: Well I think probably, so I think the problem here is going to be that — so first of all, hyper-casual is like this weird, it’s like as a market or as a sub genre within games. It’s really interesting because, you know, they basically just buy and sell to each other. It’s like…
John Koetsier: Yes, haha, they need a user for a week.
Eric Seufert: Right, exactly. So it’s just like this massive traffic trade that’s happening. It’s like, you know if you go to a — well no one’s done this recently — but if you go to a baseball game or a football game and everyone does the wave, right? That’s kind of what it is, it’s just this like cyclical, you know, just like a regurgitation of users. So, I don’t know, my sense is, my instinct is that like CPMs go down because you can’t value the higher value people, right? And so you just sort of, it just sort of normalizes at some global mean which is going to be really low because you’re not able to segment out the high value users.
John Koetsier: Well, and CPMs are already low for hyper-casuals because you don’t make huge amounts of money. You don’t have significant long term users as well in hyper-casual in most cases. What it might actually do as well is it might further incentivize the hyper-casuals to have more and more of their own games and recycled users among their own. And if they are in their own apps, then theoretically, they could target them in their own owned apps as well.
Eric Seufert: Well, yes, because there’s an identifier called IDFV in iOS, that’s ID for vendors. And so that is, you can use that, it’s an identifier for user level for a specific publisher’s apps on a user’s device. So you could say, hey, I’ve got my IDFV of XYZ, or whatever, this is a high value user. I see them and I can attach money within my own games, I can see what their values are and use that as my ID. I can then start cross-promoting them within my own apps.
John Koetsier: Guess what we’ve just discovered, because one of my questions for you was going to be, okay, we’ve talked about what this kills or what this hurts, or what this changes perhaps negatively, and I was going to ask what does this, what opportunities does this open?
And guess what, being a game publisher with a ton of titles across a lot of different genres just became really, really interesting. And we might see a significant consolidation of the industry where I’m going to go in with a game publisher and they’re going to have 100 titles, 200 titles, thousands of titles, because then they’ve got their own ecosystem within which they can target people.
Eric Seufert: Yeah. And I think that’s not only true for gaming, but it could be true for a lot of different verticals. So I had this theory a while ago and it never really played out, but remember Facebook launched Instant Apps, which is within Facebook, you know, you had a tab and you could click on it and there were these instant apps, you click on them and they instantly load and they’re all hyper-casual games, right. And Facebook basically became a games platform and Apple never pushed back. And I had this theory that Apple was, they were going to go to war over this.
John Koetsier: Yes.
Eric Seufert: And they kind of have in a way, because Apple let that happen, but Facebook’s been trying to submit this games platform as a standalone app for a long time and Apple keeps rejecting it. But I had a theory that other companies were going to look at that and say, hey, I could be a platform. Now I can’t have a standalone app in the App Store. I can’t say like, hey, I’m Uber, I’m gonna make a transportation platform and aggregate a bunch of different transportation related apps and then release my own Uber platform app within the App Store.
But what they could do is they could say, hey, you know, hey app developer, your app has to be an instant loading HTML5 app, and then we’ll create like a little platform section within Uber. And, you know, which they kind of have, right, because you can order a Lime bike or whatever, no not the Lime, the Jump bike — you can order a Lime I think in Uber. They have kind of created a transport platform, it just, it’s not presented as such.
But then what if they said, hey, you know, Uber, we’re buying Hopper. Uber’s buying Hopper. And so now you could have this whole integrated experience in Uber, but it’s all instant apps. It’s not a platform… anyway, so now what happens in this case, it’s like, well, okay, I’m some travel app and I’m Uber and I’m saying, hey, like, you know you’re not gonna be able to do UA, we could do it for you. We could promote you from within our own apps if we buy you and we publish your app… so I mean, I think that could create the opportunity for that to become viable.
John Koetsier: I think so as well. And that is super, super interesting also. There’s another question that came up here and that relates to you may not get attributions at all if you’re below a certain level, correct? I mean, you need to, Apple’s not going to send you notifications, it’s in the documentation if there’s like two, three, five, it doesn’t give a number, but it has to hit a certain number to be relevant. Somebody has interpreted that as censoring, I’m not sure if that’s the right word, but that might mean, you know, unless I’m getting thousands, I may not get attribution.
Eric Seufert: Yeah. I mean, I don’t think, I don’t know how to interpret that. That was pretty vague. I think we’ll get more info as it comes out. I don’t know if you’d call it censorship it’s probably just like…
John Koetsier: I agree.
Eric Seufert: Like an opportunity or just a value issue, right? But that was one of the weirder takeaways from the documentation that wasn’t really explained or quantified, like what’s the number, right? Like tell me, you know, you can’t just withhold that information from people.
John Koetsier: Exactly. Well, we’ll have to see, I mean it is early. Somebody else had asked other questions about is SKAdNetwork ready for prime time? I mean, honestly, ready or not, it’s going to have to be, and frankly we’ve seen challenges in the past from Apple with Apple Search Ads and other things like that, where they released it and guess what, it wasn’t really ready for prime time.
Eric Seufert: Yeah.
John Koetsier: So there could be some interesting, this was Mike Brooks actually, it could be some interesting times in the next few months, or at least starting in September. What I will say, Mike, is that I know, and I do some consulting for Singular, just so everybody’s aware of potential conflict of interest, they have released today a GitHub repo with code for SKAdNetwork for ad networks, for publishers, and for advertisers. So that is available right now in GitHub, it’s public and available, and I believe they’ll be announcing that shortly. So that is a possibility as well.
Eric Seufert: Yeah I think it’s just the problem with Apple is like, they’re such an awesome, they’re the most consumer centric, design first, comfort first, usability first company on the planet for consumers.
They really suck at the enterprise, and then it just historically that’s been especially true for mobile, right? Like if you think about iTunes Analytics, come on, you know, and like can they build an API? I mean, how big is, I’ve heard the iTunes development team is tiny, right. And so, I mean, I don’t know how are they going to even make this data available to people? Are they going to have an API? What’s the interface going to look like? You know, I don’t know, my guess is that it’s going to be pretty clunky when it launches.
John Koetsier: I would not be shocked. And somebody says here, Cameron says, “Does more ad spend go to Android from September?” You know, what’s your thought of that?
Eric Seufert: Well, my sense is that, so Apple and Google are usually in lockstep around these kinds of decisions. So I don’t know if it’s by September, Google has launched its SKAdNetwork equivalent, but it’s coming. I mean, there’s just no way they can’t. So I would imagine that this isn’t a harbinger of like long term budget shift to Android. In the short term, maybe.
John Koetsier: Yeah.
Eric Seufert: But I don’t think that’s going to be like a long term result of this.
John Koetsier: Yeah, yeah. Interesting. It’s been an interesting chat. We live in interesting times, and yeah, a lot will change in mobile marketing in the next few months. You’ve talked about mixed modeling, you’ve talked about going back to the future in some senses on how you evaluate your ad spend, and your marketing results. I think we’re going to be learning a lot until September comes and even after that.
Eric Seufert: I think it’s going to be awesome. I mean, what an opportunity to just totally start over, right? And learn new skills and have fun exploring like a totally new frontier here. I’m super enthusiastic about this. I’m super excited about the future.
John Koetsier: Excellent, excellent. Well, thank you so much for joining us on TechFirst.
Eric Seufert: Cheers, John, have a good day.
John Koetsier: Awesome. And for everybody else, thank you for joining us on TechFirst as well. My name is John Koetsier, I appreciate you being along for the ride.
Whatever platform you’re watching on, hey, like, subscribe, share, comment, all the above. If you’re on the podcast later on, you like it, please rate it, review it, that’d be a massive help.
Until next time, this is John Koetsier with TechFirst.