Apple Pay: a lesson in business models, big data and privacy

Apple emphasised this message recently during their Apple Pay announcement[^1]:

We are not in the business of collecting your data.

They emphasised that their business model does not rely on collecting user data. This was clearly a dig at Google.

Whilst Apple is hoping to run every purchase you make through their Apple Pay payment system, they are not going to collect that data.

In this age of ad-supported business models, it’s surprising to hear a technology business stating they don’t want your data.

When I heard that part, I could picture execs at Amazon and Google crying inside at the lost opportunity of getting some sweet sweet data. Much in the same way I can picture Apple and Microsoft execs cyring inside when Google gives away Android, Chrome OS, GMail, Google Docs, etc, etc away for “free”.

Apple pushed out this messsage over and over again. All the Apple PR culminated with this open letter and their new privacy policy. The open letter is very reminiscent of Steve Jobs’ Thoughts on Flash[^2] and Music[^3] open letters. This means it’s a statement of intention. It’s an important message about the future, strategtic direction of Apple’s business.

I don’t think this is about Apple being concerned about your privacy (maybe they are, maybe they’re not). This is about Apple doubling down on its business model of selling you shiny devices.

I think this is interesting because it’s Apple making a bet against big data. A bet that ad-supported, free products and services are ultimately going to be rejected by customers.

For a few years now, people have questioned if Apple’s traditional business model (of selling you stuff at a nice mark-up) could survive in the new world of Google, Facebook and Amazon with their heavily subsidised and free products.

I suppose this is Apple responding.

Big data, advertising and ad-supported products/services.

Businesses like Google and Facebook have huge investments in big data. For example, Facebook’s data storage 6 months ago totalled 300 petabytes (15 zeroes) and ingesting data at a rate of 600 terabytes per day[^4]. At this scale, they are literally pushing the boundaries of computer science. You can bet that’s expensive both in terms of their people’s skill in writing software and the hardware/data centres they have to buy/build.

These companies believe that contained within these gigantic data sets, there is special (magical) information that can only be learned by collecting data sets of that size. This is at the heart of their business model of selling targeted ads.

This busines model was long ago coined by the nerds of the internetz as “you are the product”. Interestingly, Tim Cook has used this well known phrase in his letter (emphasis mine):

A few years ago, users of Internet services began to realize that when an online service is free, you’re not the customer. You’re the product.

This is true. When you use Google, you are not the customer. Google Search is not the product. A customer is the person paying the business to produce a product or service. You (your data) is sold to advertisers who are the ones that pay Google (and Facebook and Twitter and etc etc). By definition you are the product.

Big data is an advertiser’s wet dream. By collecting huge data sets on you through social media and Google searches, it is possible to not only know what demographic you are part of, but what you are thinking, feeling and even where you will be at a future date with fairly high degrees of accuracy. Research like this is commonplace. Researchers recently pushed the boundaries further by finding out if your emotional state (mood) can spread to your friends via Facebook posts. This can lead to not only influencing you (just like they do with traditional forms of advertising) but influencing you and your friends…at the same time.

Understanding and collecting big data is so important to these businesses, that they are even willing to sacrifice user experience for the sake of it (as Facebook did by allowing experiements on almost 700,000 of their users without their knowledge[^5].

Apple’s bet against big data.

Big data is not just useful for the typical free-to-use-but-pay-with-your-data service/product. It’s useful, to various degrees, to a range of businesses. It can tell the coffee shop that on Fridays people go to work later and hence the rush hour crowd should be prepared for a little later. It can tell your supermarket to stock more chips and beer on a Friday night. It can tell Apple that 30% of iPhone buyers will also buy an Apple branded case if given the choice and hence they should make and sell them at a 38% profit margin[^6].

This is why Apple’s recent move to consciously know less about you - to not collect data - is so interesting. They are saying that they don’t believe big data’s payoff (it’s insights into you as a consumer and all the magical, mind-control future possibilities) is ever going to crystallise and even if it does, it’s not going to be worth the investment[^7].

Tim Cook is very explicit about Apple’s business model:

Our business model is very straightforward: We sell great products. We don’t build a profile based on your email content or web browsing habits to sell to advertisers. We don’t “monetize” the information you store on your iPhone or in iCloud. And we don’t read your email or your messages to get information to market to you. Our software and services are designed to make our devices better. Plain and simple.

I have been thinking about this and I came up with 3 reasons for why Apple might be doing this:

1. Apple has never really paid much attention to what users want. They like to tell you what you want and take away everything they don’t think you need. This is at the core of Apple’s product design philosophy[^8].Therefore, having insight into who you are and what you like is not really all that important to Apple.

2. This offers Apple a huge point of differentiation from Google (whom Tim Cook pointed to as their main competitor)[^9]. Under Google’s current ad-supported business model, they simply can’t take this approach or they will go broke.This makes it a risky move for Apple because it may turn out that people don’t really care about their privacy when in return they get a great product or service for free or heavily subsidised.

3. Big data insights aren’t all that accurate (and possibly never will be). When I watch YouTube I see a lot of ads for Toyota.Yet when I was buying my car:

   • I researched Ford car models using Google Search.
   • I emailed Ford dealers for quotes via GMail.
   • I stored brochures for Ford cars on Google Drive.

Yes Toyota, your money was well spent on Google. They targeted your product perfectly!

The thing with big data, and the entire digital media industry that surrounds it, is nobody knows if it works. There are lots of examples of it working. And there are examples like mine above that show that it doesn’t work all that well (or maybe it does and my next car will be a Toyota). The beauty of it is, if it’s not working, that’s because they need more data; and if it works, that’s because they told you it would work. The typical rainmaker argument - you can’t lose until people stop believeing.

Understanding business models is key.

From reading all this, you may think I have a bug up my arse about privacy and big data.

Not at all.

I see these topics from the point of view of scientific curiosity[^10].

I find this all very interesting when looking at it from the point of view of the business model that drives these decisions. When you cut through all the spin, you start to have a glimpse at what this is all about.

Benedict Evans says it best here (@BenedictEvans) 25/09/2014 7:55 am:

Apple does not collect data in much the same that Google does not charge. They’re both giving someone else’s business away for free.

This tweet sums it all up nicely. Understanding the business model helps you understand why these businesses are doing what they do.

It explains why Android phones and Chrome laptops are cheaper (no licencing fees for the OS in exchange for user data); why Facebook will pay $19 Bn for Whatsapp (more users == more data); and why Apple bought Beats and hired top people from Yves Saint Laurent and Burberry (they need people who know how to make and can influence you to buy shiny things).

It’s still not clear to me at this stage if either business model is better than the other. They both work well. Both business models incentivise the business to make better products and make the user experience better. Google Search is a terrific product. So is a MacBook Pro. It’s just very interesting that Apple, the biggest player in the tech industry, has just said no to one of these business models.

I intend to write more posts like this sharing some of the things I’ve learned about the tech industry.

[^1]: Eddy Cue at 50min, Apple Event September 2014.

[^2]: Archived link in case it goes down .

[^3]: The original version of this has been taken down. My archive link.

[^4]: Facebook’s Engineering Blog: “Scaling the Facebook data warehouse to 300 PB”

[^5]: The Guardian: “Privacy watchdog files complaint over Facebook emotion experiment”

[^6]: These are all made up examples. But you get the point.

[^7]: All that data has to be stored somewhere and is hugely expensive in terms of Computer Science expertise and hardware. See the Facebook example above.

[^8]: How’s that USB port on your iPad going? Or that removable battery on your iPhone?

[^9]: “But when I think of competitor, I would think of Google.” - Tim Cook on Charlie Rose Season 23 Episode 9 Aired on 09/12/2014.

[^10]: Probably because I’ve concluded we don’t have much control over this. Hi NSA.