>Any sales generated by the ads would not have existed without the ad, and so it's not taking anything from the app developer.

This argument is not very convincing in my opinion, as it looks like Apple is purchasing ads for brand keywords. If the ad wasn't there, the customer would go to the brands website, which would be ranked #1 without the ad.

Without the Apple's ad, wouldn't there be some _other_ ad instead? Presumably a competitor's, or the brand's own, paying the Google tax?

Yes, but if their own ad converts the user to a paying customer, they don't pay 30% of their revenue to Apple.

Yes, but an ad to a competing product would become very expensive, as it wouldn’t convert and it’s cost (cpc) would increase as it is deemed less relevant.

Or the the developer would run their own ad.

But those are very different propositions to a middle man running a ‘genuine’ ad which simply takes a cut.

The example shows the brand name being ranked 1st. For this to be a poor result for the brand it would require cannibalisation rates to be high enough to push the 15/30% fee higher than their margin - that’s not realistic because the app store price would never be at a loss to begin with (with many alreadying increasing the app store fee to match the 15/30% cut.) In all likelihood the end result is a net profit for the developer at the expense of their competitors.

Also neglected for consideration: the developer doesn’t pay for these ads - and the scarcity isn’t high enough to meaningfully affect bid pricing over regular competition.

It doesn't need cannibalisation rates to be high enough to push the fee higher than their margin to be worse off, it just needs (delta between direct and App Store margin) * cannibalisation to be greater than the opportunity cost of users who would have given up trying to subscribe to HBO if they hadn't seen Apple's search result [assuming click costs/volume differences roughly even out]. I don't think it's exactly difficult for people to find a link to HBO Max from those results if the one that says "Download the HBO App" isn't there

Sure, they still make profit from the Apple signup, but less than they would have done if people were signing up from their website.

Needless to say if the App Store price is increased 1:1 to match so the developer definitely doesn't lose, it's a terrible deal for the consumer instead (if you're Googling, it's not even like Apple can claim the App Store is adding value with simplicity of discovery...)

I run a big portfolio of mobile apps, and the discussion around this is misguided. It's not about Apple taking 30% of the revenue generated from these ads.

It's about Apple driving up the user acquisition costs for these companies so much so that it become entirely uneconomical for them to buy ads that direct users to their own websites, and instead the campaigns that target users to download the app - which results in 30%-to-Apple IAP subscriptions - become much more attractive again. So Apple is trying to make the cost of running these ads so prohibitive to the companies that they stop trying to drive web subscriptions and instead go back to driving app subscriptions only, where Apple gets 30% of everything.

As you can see it's even more sinister than it first appears - it's not a short term land grab, it's a long term strategy to prevent developers from legitimately acquiring users outside of Apple’s walled garden.

(Looking at the economics make this even more clear. Let's say Tinder has a $100 subscriber LTV (lifetime value). If the user purchases the subscription in the app, Apple takes $30 of that, so if Tinder wants to run a marketing campaign on Google, Facebook, etc that drives an app install, they can't pay more than $70, otherwise their spend has a negative margin.

On the other hand, if tinder can use these ads to get people to subscribe on the Tinder website, they have a ~$97 LTV ($100 minus 3% payment processing fee via stripe/adyen/etc). So now they can run a campaign on Google where they can spend up to $97 to acquire a user, much more than the $70 before. And because Google and Facebook inventory availability scales non-linearly with your maximum bid, a 38% increase in acquisition cost allowable could mean a 100% increase in available inventory, and potentially higher quality inventory at that.

But if Apple - with their unlimited cash pile and not caring about negative margins - comes in and soaks up all this inventory by bidding the same $97 for every user, all of a sudden the cost for Tinder to acquire these users goes way up and becomes negative margin. At that point, the rationale thing for tinder to do is stop running these campaigns. This means they stop getting web subscriptions, stop diversifying their business away from Apple, and Apple maintains its iron grip on Tinder.

Remember, in this case Apple is paying $97 to acquire a user who will generate $30 for them (30% of the $100 LTV), so they're massively in the hole on this spend. But they don't care because their goal isn't to profitably acquire users; their goal is to make the costs for Tinder to create a more diversified business so high that Tinder stops trying to. That's some f-ed up sh*t.

I don’t entire understand the ad market, so sorry if this is a stupid question, but if Apple bids $97 per user and wins, then that’s more than the $70 per user Tinder would get, that makes sense. But since that user didn’t cost Tinder anything, aren’t they +$70 since Apple ate the acquisition cost, instead of +$27 where Tinder only had to spend $70 (under the no-Apple bid price) to earn $97?

You’re 100% right - don’t doubt your logic! :)

The problem is that what you describe is the short term positive impact to Tinder. But that’s far outweighed by the long term negative of having your whole business dependent on the whims of Apple. For example, anytime Tinder wants to make an app update, Apple has to approve it. Well, if Tinder goes after Apple with regulators by filing complaints and pointing to antitrust violations, Apple now has the power to mess with Tinders business. Or if Tinder wanted to offer its users a 20% discount as an inducement to subscribe directly with Tinder rather than through Apple, Apple can and does prevent them from doing that right now by refusing to allow Tinder (or any developer) to reference in the app the fact that lower prices can be achieved by subscribing via web. The examples are countless, but essentially the more reliant Tinder is on Apple, the worse for Tinder and the better for Apple.

> You’re 100% right - don’t doubt your logic! :)

Thanks ;-). I wasn’t sure there wasn’t something I was missing.

> the more reliant Tinder is on Apple, the worse for Tinder and the better for Apple.

So you’re saying that the long-term impact for going through Apple IAPs isn’t just the 30% margin, but also increased reliance on Apple, and less knowledge about their customer base? That makes sense, I get why Tinder wouldn’t want that. Thank you :-)

Sorry, where in the article do you see that Apple is running these ads in a way that loses them money?

The article doesn't have to say it because it's just the way that google ads work. It's a bidding-based system, so at the most basic level the person who bids the most wins the auction and their ad is shown. If the LTV of a customer is $100, and there's a 3% processing fee for web transactions, then Tinder will bid up to $97 to acquire a customer. So Apple has to pay at least that much to acquire that same customer in order to win the ad auction. Once Applr has paid that $97, the user now downloads the app and subscribes via Apple’s IAP system. Apple takes 30% of those revenue while Tinder gets 70%; so Apple gets 30%*$97 = $29. So they've spent $97 and received $29 ----> they have a negative $68 margin on that spend.

I agree that that's one way it could work, but it seems really unlikely to me that Apple would be willing to lose so much money on a project like this. Some other possibilities that I think are more likely:

* Apple has a higher LTV estimate than Tinder for this traffic.

* Tinder has less available capital, so they are not able to outbid Apple even though they think that they would still earn lots of money at that price.

* Perhaps someone who is looking to subscribe to Tinder is likely to succeed even without advertising, but without Apple's ads it won't be via Apple's IAP. The amount it's worth to Tinder for a subscriber to come to them directly instead of via Apple and the amount it's worth to Apple for a subscriber to come in via IAP are about the same.

(Disclosure: I work on ads at Google, but not search ads. Speaking only for myself)

Remember, it's not that much money for Apple especially considering their whole app store business model depends on taking their 30% tax. Maybe we're talking about $50m/yr vs app store revenues of $10b (?) a year (and 70%+ margins per docs in recent court cases). They would absolutely be willing to lose this money if it extended their stranglehold on app developers - it's the same reason they fight tooth and nail in every jurisdiction around the world to prevent regulation of their app store behaviors and fees (Japan, Korea, Netherlands, UK, Australia, Arizona, US federal, etc). These lawyers probably cost them about $500m/yr (without revealing my identity trust me that thats a very reasonable estimate).

As to your bulletes points:

- Tinder is owned by Match Group who - before Tinder - spent 20 years building a paid acqusition machine. In order to do paid acquisition you have to deeply understand the LTV of your users. That methodology, refined iver years at Match was ported to Tinder (just read Matchs earnings calls). While Apple has access to all ybe transaction data of apps on iOS, so do then defelopers, who are highly resourced and highly motivated to understand their LTV/CAC. So no, I dont believe for a second that Applr has an advantage here. And even if they did, applr only collects 30% of the revenues - how could they ever guy profitable when bidding for the same slots as the developers who Are getting 70%?

- Capital - nope. Match produced close to a billion dollars a year in cash flow. HBO billions. Capital isn't an issue for either of them.

- I disproves this hypothesis with the LTV illustration above. To be clear: Theres no scenario where apple can be profitable on this spend when they can only ever get 30% of what the consumer spends.

Sorry, I don't understand how your response applies to my third point? Tinder would not be willing to spend $97/user because that is only worth it for users they would not otherwise acquire. In this case, I'm positing that these are users who already want to subscribe to Tinder, and are going to do it somewhere, the only question is whether Apple can get them to sign up through the IAP flow and take a 30% cut. If I'm thinking about this right, this means that it is worth just as much to Apple to get one of these users to sign up through IAP as it is worth to Tinder to get one of these users to sign up directly?

Ok maybe i’m not following your question then - do you mind rephrasing and asking again?

Sure! Imagine we're talking about users searching in a way that indicates very strong intent to subscribe, like "how do I subscribe to Tinder". These users are almost certain to subscribe. There are two ways that can happen:

a) They can click through to a Tinder site, where they subscribe directly.

b) They can click through to the App Store, where they end up subscribing via Apple's in-app purchases. Apple gets a 30% cut.

Assuming your $100 LTV from before, in (a) Tinder makes $97 and Apple makes $0 while in (b) Tinder makes $70 and Apple makes $27. Tinder clearly prefers (a) while Apple prefers (b), but by how much?

Tinder: they make $97 - $70 = $27 more in (a)

Apple: they make $27 - 0 = $27 more in (b)

This means both companies are willing to bid approximately the same amount, since their profit on winning, relative to what would've happened otherwise, is $27.

So are you saying that the clearing price for the ad inventory they’re competing over is $27? (If so I’ll explain why that’s not the case)

Yes: I would expect them each to be willing to bid up to $27

I would expect Tinder might be willing to bid more than $27 (per conversion) even if they only ever used Apple payments.

Any marginal conversion that Tinder creates (either against a competitor or non-consumption) is worth over $50 net to them using Apple payments. ($50 would assume a generous $20 marginal cost to serve.)

That doesn’t mean that the data that Tinder’s ROAS team is looking at will guide them to spend more than $27 in this case, only that it might.

Likewise Apple could spend $27 on every customer they converted who was already going to be paying Tinder on the web to buying via the App Store plus $27 for every user they added to Tinder who paid on the App Store who would have not bought or would have gone to a competitor (minus any credit Apple would have earned there) plus an amount of cashflow and enterprise value created by people seeing ads for Apple’s store and concluding that Apple has more/better apps than their competitors.

If we confine only to first-order effects and assume advertising is exclusively about changing payment preferences of users who would convert anyway, I’d expect both companies to be willing to bid close to $27.

> Theres no scenario where apple can be profitable on this spend when they can only ever get 30% of what the consumer spends.

I mean, maybe if tinder's margin was less than apples, but that's pretty unlikely.

Think about this way: both companies are bidding on the same keywords. One party (the developer) collects either $97 if the consumer signs up on their website directly or $70 if the consumer downloads the app and signs up via the app store. The other party (apple) collects only $30 (at a maximum) if they drive the user to sign up via the app store (30% of a $100 LTV).

If they’re both bidding on the same keywords, which they are, then the price of the keywords is the same to each of them. It’s an auction, so whoever bids the most will win the auction. The developer can spend up to $97 to bid on that keyword and still be profitable or at worst break even, while apple can only spend up to $30 to stay profitable or break even. So if Apple is winning auctions, it means they’re spending more than the developer, despite having less than half the purchasing power of the developer. So they’re spending negatively almost no matter what.

Both ads are shown though with Apple in the 2nd slot. So HBO is bidding more and I doubt Apple is going into negative margins.

Again, by definition theyre going into negative margins because they can only ever get 30% of the revenues; how could they compete with the developer who gets 70% of the revenues without going negative?

They're not beating the developer. They're getting the #2 slot.

Remember two things:

1) more demand in auctions means higher prices. so if apple is taking the second slot they’re almost certainly increasing the cost of the 1st slot for hbo. which is apples whole point: to increase the cost (literal and figurative) of hbo’s efforts to get people to sign up via web rather than via app.

2) in the world of search marketing - including both on google and the app store - there is a strong belief by brands that they MUST be the #1 paid result for their branded search terms, no matter what. so what you’re seeing here is hbo saying “we always have to be #1 on our branded search no matter the cost”. This is exactly why people complain constantly about google showing competitors ads when the original company’s brand is specifically searched for by the user - effectively the competitor is freeloading on the incumbents brand searches to drive traffic (by paying to show ads on those searches). Google doesn’t care because they make more money this way (more competition for the branded keyword drives up CPCs).

all of this is to say that apple being in the #2 slot doesn’t say anything about the intent of their actions here, which is to make it uneconomical for companies to move users off apples rails onto their own systems.

Both sides run services which have costs, the customer acquisition needs to take into account the potential profit of having the customer and not just raw revenue.

Maybe Apple clears 15% of the user's purchase price after credit processing, tied-in platform development costs, customer service/hosting fees, and the 70% payout to HBO, and HBO clears 15% after licensing, hosting, app development costs and the 30% fee from Apple.

In this case, HBO would be motivated to have a website sign-up to raise their 15% to closer to 40%, and Apple would be motivated because otherwise they are negative for the lifetime of the subscription.

This is a second price auction, so the person with the highest bid pays the amount of the second highest bid. If Apple is bidding $97, their cost to acquire the customer may be substantially lower than $97 (their cost is whatever the second highest bid is), but they are forcing others in the auction to spend at least $97 to get the customer.

Thanks for this insight! Very interesting.

Unlike predatory pricing, you can’t really price people out of this permanently.

Tinder in this example is still getting the customer. Google is still getting paid.

So the moment Apple stops buying the ads, they’re both ready to participate again. So how is it a long term play?

You're right on two accounts: - Google gets paid no matter who pays, so they don't care - Tinder still gets the revenue (and at better margins bc they're not actually have to pay for that user anymore - Apple is paying)

And even your last point is not wrong: at some point Apple may stop doing this. But that could be year's away, and in the meantime, they're throwing their big stack around to make it too costly for developers - who Apple supposedly partners with - to build businesses that are less dependent on Apple’s whims. Plus, Apple uses the fact that most subscribers to app store products subscribe on the app store itself to bolster their case with regulators that no reform is needed bc consumers are overwhelmingly happy to use Apple’s IAP systems. But if Apple is putting it's finger on the scales in order to actively drive users away from web subscriptions, then they heavily misleading these regulators about the true “choices” consumers are making.

> how long

I think we can assume Apple is getting enough out of it to run this scheme permanently.

Unlike Tinder, they can get extra revenue from the user staying in the ecosystem (active credit card registered for IAP), getting used to Apple’s service, and keep buying Apple devices. Their LTV of the user is not just the 30% cut.

You do it while a service is growing, to capture most of the subscribers. Once growth dies down you stop buying and while the subscriber base will slowly bleed from app to web, you can ride the revenue stream from the growth fase’s subscribers for a long time.

On the Amazon part, the harm is caused by your book losing value during the Amazon give away and a while after.

For instance any other seller of your book is affected, so you potentially lose distribution channels.

Customers also get attracted to Amazon, but your book is only seen as a bait, and when it goes back at the regular price it will be less desirable (some will have negative feelings of having missed the sale, some will keep seeing your book as something only worth giving away).

If your book has a short shelf life and you didn’t expect much of it, or if it’s a stepping stone into a series and you could have given it away anyway, Amazon footing the bill is a boon. Otherwise it probably shortens the life of your book.

"A reseller offering both trade mark goods and competing products is not precluded from using the trade mark in advertising, provided that the reseller safeguards the legitimate interests of the trade mark owner. The German Federal Court has found that if a trade mark is used misleadingly in advertisements originating from a Google search resulting in the exploitation of a trade mark in a way that guides the public to third-party products, such use may provide a ground for the trade mark owner to oppose the use of the trade mark."

This was specifically re: Amazon using a trademark in AdWords to redirect sales to Amazon, where they then of course also sell competitor products.

This might be moot of course if you granted Apple rights to your trademark in whatever agreements, but then these agreements are also frequently found inapplicable so ymmv.

To me it’s an issue if Apple refuses to stop if a particular business asks them to. That seems like an easy measuring stick.

Not sure they can. The dev agreement permanently signs away their refusal.

This shows just how much a monopoly Apple has over brand <—> customer interaction.

They say that their app store not allowing side loading is for user safety.

In actuality, it forces the brand/company/startup into a place where they're completely shackled. They pay 30% taxes, have no customer relationship, and have to step through every hoop to satisfy apple.

The handicap their browser so you can't get new runtimes or deploy software via alternate means. Good luck launching your media startup on iPhone without an app.

Now this. They're clearly trying to gobble up any interest in other apps and force them into the payment shackles.

This is a fucking monopoly! Please stop this, department of justice! Apple has undue power in computing! Every company has to go through them to reach 50% of consumers, and apple is extorting us!

Apple says you're welcome to go to consumers another way, and then they do this. Look at their actions.

"nothing will fundamentally change"

-- the person currently in charge of the DOJ, when running for office, speaking to donors.

Yesterday there was some discussion of "Project Bernanke", so-called "quantitative easing" for DoubleClick's AdX. One could see similarities.

Here, Apple is helping certain "preferred" iOS app developers win against other iOS app developers. Apple stands to profit, the preferred developers stand to profit, but not other developers (and possibly users who will never discover their apps).

There, Google was helping Google Ads users, i.e., "preferred" DoubleClick AdX auction participants, to win against other participants in AdX auctions. Google profits (and brags about it in its internal presentations), Google Ad users may profit, but not the users of Google Ad alternatives.

Big Tech wants to facilitate "winner-take-all" scenarios. Healthy competition and parity amongst competitors (for user attention) reduces the profitability of the "tech" company business model. Big Tech would rather that user attention only be focused on what is most popular, creating more desirability to advertise to that single audience. A "Top 20" mobile app list, or top of page 1 of SERPs.

Concentration of web traffic, 80/20, network effects, and so on. Winner-take-all. The most ideal conditions for selling apps or ad services. The issue is not that they are doing this, it is in how are they doing it. They run and control the so-called "platforms" for advertising and sales and use that privileged position to manipulate outcomes that favour themselves. (And stifle competitors.)

Exactly and the crux is how can you prove the download would happen regardless of the ad put out by apple . This is an attribution problem on the ad space and nobody has solved it yet. Google has been trying to push complex attribution rules but they are not in a position to do it because they are insentized to attribute the sales to their own ads.