The war Microsoft should have won
Microsoft had many things in its favor to become the dominant mobile OS, but a number of issues, many home-grown, saw it lose its early start to Google and Apple.
In 2012, Vanity Fair published a long story on Microsoft under Ballmer and the “lost decade.” I’ve re-read it many times and a quote keeps jumping at me:
“You look at the Windows Phone and you can’t help but wonder, How did Microsoft squander the lead they had with the Windows CE devices? They had a great lead, they were years ahead. And they completely blew it. And they completely blew it because of the bureaucracy.” Ed McCahill
The quote speaks to me specifically because I joined Microsoft’s Windows Mobile team as an intern in 2002 and then full time in 2003 and at that time I really felt we had a chance to be one of the leaders in the upcoming evolution of mobile phones into internet-connected computing devices.
For those of you that might only remember the smartphone race as a Android vs iOS battle it is worth providing a brief summary of where we came from:
Our joint evolution towards smartphones started with Personal Digital Assistants (PDAs). Apple famously had the ill-fated Newton. Palm was the first mass-market mobile PDA. Microsoft soon launched PocketPC based on its WindowsCE code base. Symbian (made by European mobile OS company, Psion) in partnership with Nokia pushed the envelope and integrated PDA functionality with the launch of the 9210 Communicator in 2001 (important to note that Nokia had launched a mobile phone with an OS as early as 1996, the Communicator 9000, considered by many as the first “smartphone” and the 9110 in ’98 both running on GEOS).
So in 1999, when Bill Gates announced the Smartphone initiative based on the WinCE kernel, the OS world was already messy. But one could argue Microsoft had a number of things going for it to win this fight:
- A solid and stable embedded OS code base with WinCE and a growing PDA platform in PocketPC which integrated familiar apps and user experience from the deskptop
- A relationship with chip manufacturers and OEMs which should allow it to copy the model of the Wintel era onto smartphones where Microsoft provided the OS, reference designs and marketing dollars and OEMs built the hardware and took it to market
- A well managed and broad set of application developers who lived and died by Microsoft and would surely support its new shift towards a mobile platform. This also included Microsoft’s own apps like Outlook, Word, Excel, MSN Messenger, Internet Explorer and XBox assets.
- A lot of money in the bank to buy customers and market share
So where did it go wrong and how did Microsoft lose the chance to become the leading mobile OS? [NB: In between 2002–2006, I was a measly post-MBA Level-60something member of the Windows Mobile team. These insights are simply my views from having been in the trenches from Microsoft, from having been in the mobile internet space since the early days of WAP and also from having been part of Google’s mobile efforts a few years later. I am sure much more strategic and thoughtful discussions happened in the Boardroom of Building 117]
- Microsoft was fighting a battle on too many fronts: As the graph below shows, the early mobile OS wars were not only between Nokia and Microsoft but also between Palm and Canadian pager-turned-drug RIM/Blackberry. Nokia/Symbian owned the relationship (and majority of dumbphone and smartphone market share) with operators outside of the US. Palm had a cult following in the PDA space and entered the connected device with the Treo. RIM played a brilliant strategy of seeding its crackberries to banking MDs where it became a must-have device for aspirational i-bankers. To fight all these battles Windows Mobile had too be too many things to too many people. A PDA replacement as good as Palm, a phone that worked as well as Nokia phone, a connected device that synced with email as well as RIM. In the end it was too much too jam into a small box.
- Mobile OEMs were not the guys Microsoft was used to pushing around: HP, Dell, IBM were all willing co-conspirators in the Wintel cartel. But none of them had a major role in the mobile ecosystem. In the early 00’s the key mobile players were different players Nokia, Siemens, SonyEricsson, Samsung…
Having seen the marginalization of the hardware makers in the PC world many of them were hesitant to partner with Microsoft which is why Microsoft was forced to launch its first generation of smartphones with unknown “ODM” HTC. [NB: my job at Microsoft in 2003 was to partner with Samsung and Motorola, the first Tier 1 phone manufacturers, on launching the first branded smartphones, the MPx200 and the i700 in partnership with Motorola and operators around the world. Neither would be a huge initial success]. SonyEricsson would not agree to do a Windows Mobile device until many years later and Nokia…well we know how the burning platform worked out a decade and a half later.
- Enterprise, consumer, enterprise, consumer: The single biggest driver to the failure of Windows Mobile to take off, was its key asset: Microsoft. In 2000 Microsoft hired Juha Christensen from Psion (the maker of Symbian) and his call to arms by the time I joined the team was to make the smartphone battle a “two-horse race” between Nokia/Symbian vs Microsoft. This strategy, by default, meant a large phone volume play, and therefore a consumer-centric offering. This meant putting Microsoft’s consumer assets (MSN Messenger, internet explorer, XBox) front and center. But Windows Mobile was a small business inside of Microsoft, and the cash cow still lay with Office and server products. RIM’s Blackberry Enterprise Server was slowing down upgrades to new versions of Exchange and the edict was given to Kill RIM (err sorry after the DOJ case, Microsoft employees were no longer allowed to say “kill”… “RIM-compete” was the right DOJ-approved title for the strategy). And so the focus of the small division flipped to a more narrow shipment volume of prosumer/enterprise-centric devices with Outlook and productivity as its core. The consumer/enterprise flip-flop would happen a few more times in the years to come. Juha left in the summer of 2003 and the two-horse race changed from Nokia vs Microsoft (consumer) to RIM vs Microsoft (enterprise).
- Consumers drove the purchase decision: The Windows logo sold PCs. Ergo the windows logo (and all the familiarity that it implied) must also be able to sell phones… but it turns out that consumers didn’t yet care what OS the phone was running. Consumer just wanted their sexy RAZR flip-phones, or 8 mega-pixel Nokia camera phones, or low-cost LG phones. It was still, at that point, just a phone and the primary function was still primarily making a phone call (something some of the early Windows Mobile devices weren’t always that good at… and yes there was a way to CTRL-ALT+Del a WinMo device).
- One-handed versus two-handed, keyboard vs touch screen: The PocketPC PDAs had been two handed-devices with stylus as inputs. The vision for Smartphone was always for it to be a one-handed device. It was actually pretty good as a inbox triage tool with one hand… But market feedback was puzzling. PocketPC was planned to be phased out to focus on Smartphone, but two-handed input as a consumer choice was still driving sales. As the iPhone and Android devices would prove soon-enough, consumers were willing to accept two handed input if it provided for richer features and bigger screens. Microsoft, given RIM-compete as an edict, focused on keyboards over touchscreen.
- The business model looked back not forward: Microsoft had made gobsmacks of money selling licenses of Windows to OEMs. It was the Coca-Cola syrup of the tech world. Surely the same model should work on mobile: Charge OEMs a fee for the Windows license, let them take care of factories in Taiwan. But in the mobile world, the power dynamics and customer expectations were different. Operators were still subsidizing phones for consumers and the power lay with the operator more than the OEM (or the OS maker). A $10 Windows Mobile license had a massive impact on the Bill of Materials (BOM) and therefore the wholesale value of the phone. As the new Treasurer of the WinMo group once said to me “We’re thinking about it the wrong way. We need to move from one time license fee payments and move towards recurring revenues from services”…If only his voice had been heard [Xiaomi is using this same strategy aggressively pricing phone but focusing on long term value of services]. Apple upended the old model by launching a device consumers drooled over and were willing to pay $500 for (unsubsidized!). Android changed it even more by making an OS that was richer and more open than WinMo completely free. The old Microsoft business model was never innovated upon. Other innovated around Microsoft.
- The “web” was not the “web”: The final nails in the WinMo coffin came in 2007 and 2008 when Apple launched the iPhone and Google launched Android. And it was not about the design of the iPhone or the amazing hinge Andy Rubin built for the Android G1, but more philosophically about what type of web consumers wanted on their mobile device. Microsoft, RIM and Nokia had all built ways to compress and reformat the web into smaller screens. These phone and OS makers seemed to believe they had the right to determine what the web should look like on a mobile device. Android’s vision had always been to have a full rich-HTML web experience on a mobile device (very googley) and both the iPhone and Android platforms launched with webkit browsers and full HTML support. And consumers voted with their thumbs…They wanted the “web” to be the web. [NB: This notion of the mobile web was the core issue of the partnerships I would be tasked with cementing between Google and Nokia and RIM a few years later, but that’s for a story for a future post. Short version: Google was right.]
So in hindsight, the prediction that Bill Gates made in 1999 on the disruptive force that mobile would have was correct. And the bet Microsoft made on a mobile OS had a high likelihood of succeeding. But at it’s height, Microsoft never got above 15% of the Mobile OS market, and as Benedict Evans has pointed out it is now clear that Google and Apple won the mobile war.
Having been in the front lines of the early day of the war many years ago, I have had time to reflect on the various mistakes that I saw inside of Microsoft and then of Microsoft from the outside, and have come to the conclusion that it wasn’t one single error but a number of them that took away a prize that, thinking about the lay of the land in 2000–2005, could have rightfully been Microsoft’s…
And it would be disingenuous to end this post without a link to the infamous video of Ballmer laughing at the iPhone when it launched… “It doesn’t appeal to business customers because it doesn’t have a keyboard…” “Right now we’re selling millions and millions phones a year and Apple is selling zero…” “I like our strategy…I like it a lot!” A few years and billions spent buying Nokia later not sure it was as funny. This reflective Charlie Rose interview in 2014 is also worth a watch.