Apple's new iAd proposition has been generating a great deal of discussion lately, most of it positive, and most of it remarkably short-sighted. It seems most people, including Steve Jobs, have forgotten the basic lessons of computing and the internet. People who forget history are doomed to repeat it. The iAd has no future, and neither does the iPhone/iPad. I will show why iAds must inevitably die, and how Steve Job's strategy for iPhone and iPad will inevitably lead Apple into becoming at best a marginal niche player, at worst an ex-business.
We need to start by remembering what the smartphone platform is. First, let's forget the archaic concept of a "mobile phone." The iPhone is no more of a phone than a PC running Skype. The majority of iPhone time, money, and resources lies in apps. iPhones, just like smartphones running Google's Android or Microsoft's WinOS, are not phones but small computers that happen to have telephone capabilities. They may have started as phones, but they outgrew that classification a couple of years back. The only serious difference between a smartphone and a computer is size and the fact that smartphones are inherently location-aware. You will notice that the creators of smartphone operating systems are all computer companies.
As computers, smartphones are subject to the multi-layered business model common to all computers. Technology manufacturers, such as Nokia and Samsung, build the physical hardware. Above them we have the providers of operating systems, of whom the major players are Apple, Google, and Microsoft. Other companies provide apps and services. Some services are delivered directly through the operating system, such as SMS and phone calls, while others are delivered within applications themselves. Under this model, advertising falls into the category of services, while widgets combine apps with services.
Understanding the dynamics of this marketplace requires recognizing where the power lies, and how it shifts as technology evolves.
The smartphone marketplace
Hardware manufacturers make their money by selling new phones. They have no interest in making phones that last forever, or that can be upgraded via software. Just like PC manufacturers, their only chance of continuous revenue is by continually developing new models. The best way to sell a new model is to provide new capabilities, but opportunities for new capabilities are limited if the operating system does not provide access to them. Since operating systems are more difficult to develop than applications and services, and cannot upgrade as quickly, the emphasis on exploiting new capabilities inevitably falls on the app developer and service provider community. Thus the success of smartphone hardware providers is inextricably linked with the development of an active mobile app marketplace and a dynamic service provider community.
Irrespective of the capabilities of the hardware, both apps and services are limited by the capabilities of the operating system. If the operating system does not permit access to a new hardware capability, then that capability cannot be exploited. If the operating system does not provide an efficient development environment, then it may be difficult to exploit the new capability in a cost-effective manner. Thus hardware manufacturers, application developers, and service providers are all dependent on the operating system.
Operating system vendors can make money from both sides of the equation. They can license the operating system to the hardware manufacturer, and they can charge licensing for apps built on their platform, or for developer kits and support.
Operating systems are only attractive to hardware manufacturers to the degree that they offer future-proof access to new capabilities the manufacturer may one day create, and to the degree that the operating system is attractive to application developers and service providers.
It is important to remember that developers do not build applications for hardware, they build applications for operating systems. If the app works on Android, it makes no difference who built the hardware, the app will work on all Android phones.
Operating systems are attractive to application developers and service providers to the degree that they offer profitable development paths. For a developer, profitability is a combination of ease of development (which determines cost), and the deployed base of the operating system, which determines the size of the market. No one is interested in developing an application for an operating system no one uses.
The critical component in all this is the customer. In order for everyone to make money, people have to purchase the hardware, apps, and services. Smartphones, like all other computers, sell on the basis of what you can do with them. People buy IT equipment (laptop, PC, or mobile) on the basis of the applications they can run on it. The item purchased needs to be able to do what the customer wants it to. There are so many programs around for PCs today that this is rarely a consideration -- almost every conceivable application you could want is available.
As a result we have largely forgotten that capabilities are central to sales. However, there are many instances in which purchasing a Mac is not an option because the required software does not exist, which shows that applications still control purchases.
This creates feedback loops in the smartphone marketplace -- hardware is sold on the basis of the range of apps available. The range of apps available is dependent on the operating system. The success of the operating system depends on providing a good platform for apps and services. An operating system is attractive to a mobile hardware manufacturer only if they think it will help them sell hardware, and in order to sell, they must have apps and services.
So an operating system's success is dependent on being an attractive platform for developers. In order to be attractive, the operating system must have (or promise) a large installed base. Thus hardware will sell if there are apps and services. There will be apps and services if there are customers. There will be customers if the operating system is widely deployed. An operating system will be widely deployed if it sits on a popular hardware platform, or as many less-popular hardware platforms as possible, and if there are attractive apps or services for it. Everybody feeds everyone else.
Apple's success is therefore dependent on the app developer community. Since it also controls the operating system and developer access, it determines how attractive the development environment is for developers. This has always been Apple's strategy, right back to the first Mac. Unfortunately, Apple's history in this area is appalling. Apple's desire to control its marketplace has made it a poor choice for developers, even when it offers a large market. Having a large base of customers makes Apple initially attractive, but its poor support for the developer community eventually forces smaller niche players out.
The long term result is easy to see -- Macintosh now runs Microsoft Office because no one else was interested in providing a compatible office suite. Apple's restrictive policies over the Mac almost caused the death of the Apple Corporation, and it was only by opening the environment to its arch-enemy Microsoft that Apple was able to survive.
When Steve Jobs announced MS Office for the Mac to a stunned audience in 1997, he looked very uncomfortable about it (decide for yourself). He justified it by saying Apple existed in an eco-system and could not sustain the Mac as a closed platform. His iPhone strategy seems to have forgotten this painful lesson.
I first became involved with computers in the late 1970s. This was before the days of the IBM PC and MS-DOS or Windows. Computing in those days resembled the smartphone market of today -- there was no common operating system. I remember buying my first computer -- it took a week. Because each hardware vendor had their own operating system, you could only buy applications that had been specifically designed for that hardware. In order to buy a computer, you had to first identify all the applications you wanted, then cross-reference all the hardware that these apps had versions for. If you were lucky, you would find a computer that ran everything. However, it was more likely that you would find that nothing ran all your applications. You were then forced to find alternative apps until you found the best compromise.
Eventually Microsoft solved this by creating an operating system which would run on any Intel chip, inventing the concept of "PC compatibility." Now app developers have the entire PC community as a market, irrespective of who makes the hardware. At the same time hardware manufacturers can produce their equipment knowing there is an massive range of apps customers can run.
Apple never joined in the universal move to PC compatibility. Based on the Motorola chip, Apple chose to cater to niche market players with hobby computers such as the Apple II. Apple's day came later when it copied the GUI operating system being developed by Xerox and created the first Mac. The GUI posed a threat to Microsoft's survival and the dominance of the PC, until Microsoft got its own GUI right with Windows 3.0.
Microsoft's strategy was always to open its platform to the widest possible developer community, while Apple's was always to restrict and control. In many ways, Steve Jobs continued to think in terms of the world he grew up in, a pre-PC world -- each computer manufacturer producing its own operating system and strongly controlling developer access.
Apple still continues to think this way, but the success of MS-DOS and Windows have shown that it is not sustainable. At peak, the Mac had 30 percent of the small computer market. Now that share is less than 3 percent.
Mobile, cable, and web access: 1990s
When the web was developing during the 1990s, I was often involved with attempts to extend web services to mobile phone users. We never succeeded, and neither did anyone else. The reason was the mobile phone companies regarded the customers as their property, and they weren't about to let others near them. They accepted that people would want web access from their phones, but thought they could provide all the websites themselves, or license those that would be allowed in.
If you had a mobile at the time, you may recall that for the first few years web access was restricted to websites provided by the phone company.
The mobile phone companies figured they could be more than just information carriers. They thought they could be information providers as well. They didn't grasp the scale or range of demand, and thought they could restrict subscribers to the limited amount of information they could offer. Eventually, of course, they realized this was impossible, that people wanted access to everything -- all 200 million websites, not some restricted subset of a few thousand. It was inevitable that, sensing the scale of the market demand for unlimited access, a mobile company would eventually open up, so they all did. Many cable companies tried the same thing at the time, and eventually learned the same lesson.
Lessons for Apple
Both these examples show that no company can succeed in the long term by restricting customers. Such a strategy may succeed for a few years, maybe even a decade, but in the long term, the market demands open systems. Companies that understand this always win out against those that try to stay closed.
The iPhone may be popular now, but history has shown us that the days of competing operating systems eventually give way to more open platforms. The world will not tolerate three or four competing smartphone systems with roughly equal market share. Eventually, one system will dominate. Apple's iPhone OS and BlackBerry's RIM are not candidates for that role because they're not available for other phones, which only leaves Google's Android and Microsoft's WinOS as candidates for global domination.
I think the determining factor will be developer support. The more apps available, the more popular the operating system, both for hardware manufacturers and customers. Microsoft has a tremendous advantage here -- it has 30 years of experience in developer support, whereas Google has already developed a reputation for poor communications.
According to Google there are 10 times as many web pages providing developer information for WinOS than there are for Android. Microsoft also has a huge body of experienced Windows developers. Google may still win the race, but we must recognize they're starting behind Microsoft.
Right now the iPhone has a dominating position in the U.S. smartphone marketplace. However, we must recognize this is a global village. Apple cannot sustain the iPhone as a purely U.S. phenomena. However, its share of the global smartphone marketplace is small. In Q1 2010, iPhone made up 15 percent of global smartphone sales, slightly less than BlackBerry's 16 percent and much less than Nokia's 48 percent.
If the smartphone goes the way of previous computers, and the way of the internet, Apple's strategy will eventually lead to the iPhone occupying a similar niche to the Mac -- a miniscule market share sustained only by the fanatical loyalty of dedicated followers.
Apple repeats itself
Steve Jobs says he hates Adobe Flash and will not support it on iPhone. The reason is clear -- Flash provides a cross-platform development system. Build an app in Flash and it runs on every operating system that supports Flash. In this sense Flash becomes an operating system to sit above the different operating systems deployed, providing developers with the widest possible market.
If all phone apps are written in Flash, then I can switch phones without inconvenience -- I'll keep my apps. People -- both consumers and developers -- are no longer locked into Apple. Apple has shown it likes to own the entire space around its devices, but history has shown that consumers like open systems, and that they inevitably get what they want. Locking Flash out is unsustainable if you want to retain market share.
Irrespective of the merits or problems with Flash, the only potential alternative on the horizon is HTML 5. Those of you who have seen iAd demos will recall they are produced in HTML 5. HTML 5 offers solutions for deploying complex media previously only possible with Flash. HTML 5 is the route Steve Jobs recommends, and he's already sending Apple down that road. However, HTML 5 is not finished. It is not even partially complete. Currently HTML 5 is in draft. This means new features can (and will) be added, while existing features may be changed or even dropped.
It is likely to be another year or two before HTML 5 is even finalized. Once it has been finalized, it then enters "draft recommendation" stage. At this stage anyone can evaluate it and propose changes. They will do so. No version of HTML has ever gone through the draft recommendation stage without being changed. HTML 3.0 was so radically altered at the draft recommendation stage, the final version agreed upon was numbered HTML 3.1.
Once again, Apple has shown it can't learn the lessons of history. Those of you who, like me, were developing websites in the early to mid-1990s will remember the days of "browser compatibility" problems. As companies like Netscape and Microsoft battled for domination of the web browser, they would fight to support the next version of HTML before anyone else. This inevitably led to support for unfinished (and competing) versions of HTML, so that when the latest HTML standard was finished, no browser would support it properly, and each would break the standard in different ways.
Building browser support for HTML 5 at this time means the browser will need upgrading when HTML 5 is released, and most HTML 5 apps built now will be obsolete. Irrespective of how attractive it may be one day, HTML 5 is not a viable development platform yet.
Just as it did with the Mac, Apple's treatment of iPhone/iPad developers is already turning some away. The web abounds with complaints from developers who have had their apps dropped from the iPhone App Store without reason, or who have been disadvantaged by sudden changes in Apple's T&Cs. Apple's attitude to developers looks to me as if Apple feels it is doing developers a favor by allowing them the privilege of access to their customers. I can't help contrasting this with Microsoft's attitude that developers are customers, very important ones who determine Microsoft's success.
When I look at the lessons of history, Apple's own past, and how things work out, it seems to me inevitable that within 5-10 years the iPhone will hold around 5 percent of the smartphone market at best.
iAd is just a second-rate widget
Calling iAd creations "advertisements" is misleading. iAd advertisements are, in reality, widgets. I presume Apple doesn't want to call them widgets because the term feels a little sour -- we've learned that effective marketing widgets are hard to find. Would anyone be so excited if Apple had said it supported a new widget technology? You can do wonderful things with iAd, just as you can with any widget, but consumer responses have shown good widget penetration is extremely hard to achieve. Widgets that succeed provide services, so the iAd isn't really an advertising system so much as a system for sponsored service delivery.
Widgets have their place, but we already know they can't replace all other forms of online marketing. How many times have you asked yourself : "Should I deploy a banner ad or a widget here?" -- that's how often iAd will provide a viable alternative to banner advertising.
iAds can be created in Objective-C and/or HTML 5. Objective-C is a difficult and cumbersome language. It runs at one-third the speed of C++, and has a poor reputation with C developers. A number of development agencies have announced they will not produce iPhone apps because Objective-C is not a cost-effective development platform. Unfortunately, HTML 5 is merely an idea, not a reality. Keeping HTML 5 iAd widgets alive over the next few years is almost certainly going to require frequent re-coding. As HTML 5 is developed Apple will need to change browsers to support each step. Consumers will not upgrade at a consistent rate, so this will lead to the development of different versions of the same iAd widget for each permutation of HTML 5 deployed by Apple. This will significantly increase the cost of development and reduce the potential ROI. Thus, from a developer perspective, iAd is not an attractive platform.
Apple dictates -- you obey
When he announced the iAd, Steve Jobs said he didn't want advertisers to segment the iPhone audience. He didn't want behavioral targeting, demographic analysis, or other forms of segmentation. He said he wanted advertisers to treat the community of iPhone users as a single demographic. All the advertising people I have spoken to don't see this as realistic.
As Ian Wolfman, CMO of IMC2, pointed out, simply selecting the apps you'll put iAd widgets into is already segmenting to some degree. However, in the longer term, he doesn't see how agencies can make convincing arguments to clients for iAd campaigns if they can't segment their deliveries.
In order to help control this space, and prevent segmentation, Apple will not allow any third party measurement of iAds. Apple has announced it will provide 15 standardized metrics, which it will gather and report for you. If those metrics don't suit your needs -- too bad. This also means you have to trust Apple when it takes its 40 percent cut of iAd revenue. Apple measures, it charges, you trust.
As Wolfman says, "Apple will have to change. It's not sustainable, people will need to make their own assessments. I think Apple wanted to start with something simple and controlled, but I expect them to change the program very quickly."
Roy de Souza is CEO of Zedo, one of the world's leading ad technology companies. Ad metrics are core to his business. According to him, "It is important for advertisers to track with their own technology because they need their own data if they are to trust that the ads were shown to users as promised. This accountability creates a strong incentive for Apple to make sure that the ads are served correctly. Google initially did not allow advertisers to use their own tracking but now they do. I would urge Apple to follow suit."
The iAd is a symptom of Apple's inability to come to terms with the way computing has been for the last 30 years. While designing innovative products, as a business Apple still strategizes like it's the 1970s -- trying to create isolated ecosystems when everyone else knows the world wants one big open inter-connected system.
Apple seems wedded to the idea that it can own all aspects of its customer experience, even though its own corporate history shows this is unsustainable. The smartphone environment is a mirror of the early days of personal computing, yet Apple shows no sign of having learned from this experience.
The iAd system depends on unfinished technology, and therefore cannot be sustained in its current form, increasing the cost of ownership for those who develop iAd widgets. At the same time, Apple wants advertisers to forget about demographics and segmentation, rolling back the clock to the days of TV and radio-style mass delivery. In addition to asking advertisers to work with flaky technology in an old-fashioned manner, Apple proposes to deny them the ability to assess their own work, or to check that it is doing what it's being paid for. Under these circumstances, iAd only has a future while there is no alternative. History has shown us an alternative will inevitably develop.
I'll leave the final word to Ian Wolfman:
"I'm not convinced iPhone/iPad has a long-term future. It's a closed system. It's attractive now because the U.S. lacks an open alternative, but it's inevitable that one will develop."