Back in 2011, Nokia CEO Stephen Elop shook the tech world with his now-infamous “burning platform” memo. It described Nokia as a man standing on an oil rig engulfed in flames - forced to jump into icy waters below because staying put meant certain death.

The metaphor was meant to illustrate the urgency of making a decision at Nokia. The company's reliance on its own Symbian platform was no longer viable in a world dominated by Apple and Android. Elop's conclusion was a radical one: Nokia would abandon its software platform and adopt Microsoft’s Windows Phone OS instead.

We all know today that the jump didn’t save Nokia. Microsoft ended up buying the mobile phone business from Nokia and shutting it down shortly after. While the intentions behind the memo were good, the way everything was orchestrated to lead employees and customers toward a new platform left a lot to be desired.

By publicly declaring Symbian a failure before Windows Phone was even close to challenging Android and iOS, Nokia effectively vaporized its existing platform business overnight. Developers that worked on Symbian were hardly thrilled. Carriers didn’t have confidence in a third big player emerging from the alliance of “two turkeys,” as described by Google’s VP.

The message might have been strategically correct, but the execution was catastrophic: Nokia burned its ships before the rescue boat had even left the harbor. Microsoft survived but the scars from losing the battle for mobile computing must run deep.

“Burn the ships” memo of Satya

14 years later, we are seeing something similar taking place. Only this time, it is Microsoft themselves who have decided to set their own ships on fire. Not because they wouldn’t be a leading provider of cloud applications and platforms to business customers all over the world. It’s because their CEO is seeing a shift ahead that he wants to take full advantage of. Satya Nadella is determined to “make Google dance” and he wants to be the one to build the iPhones of the next decade.

Bloomberg published an article last week with the title: “Microsoft’s CEO on How AI Will Remake Every Company, Including His”. (Alternate link here.) There’s plenty of interesting insights to describe what is happening at Microsoft and why. Being deep in the Power Platform ecosystem myself, I paid special attention to what was said by Charles Lamanna who runs the Business Applications and Platforms (BAP) Business and Industry Copilot (BIC) organization:

The leader who first launched Power Apps and Power Automate in 2015, then helped Power Platform and Dynamics 365 become one thing until inheriting the entire BizApps business from James Phillips, is now being told this by his superior: “The last five years we spent building, it doesn’t matter. It’s not worth anything anymore. Burn the ships.”

Stop for a minute and think about this. What are we seeing take place here? What is the “it” that’s not worth anything anymore?

To understand the scale at which Microsoft operates, we need to remind ourselves that while the Dynamics 365 business makes billions of dollars in revenue annually, it’s not worth all that much in the end. The estimated share of revenue coming from the BizApps biz is less than 3%. It’s much smaller than Bing, and less than half of what LinkedIn brings in as revenue. See the pie chart here for more details:

Does that D365 figure include the entire Power Platform? Who knows how the categories are defined, given how the product names and organization change at MS. As an example, even though Power BI is firmly on the Fabric side now, I bet it still gets counted into the active user stats. In terms of monthly active users of Power Platform products, the latest figures are:

As usual, the numbers are big. They always go up. There’s endless excitement in the air when talking about the future opportunities for any technology that MS is selling.

But nothing is sacred. In the very first email that Satya sent out as the newly appointed CEO of Microsoft, he stated that “our industry does not respect tradition — it only respects innovation.” When you look at it from a high enough level, there’s very little that could not be sacrificed in the name of innovation when it comes to software business and the big tech corporations.

How many people have been using MS Office for decades? Well, guess what, its brand was replaced with Microsoft 365, just like that. Then, it all got bundled in with AI as Microsoft 365 Copilot. Consumer customers already got the forced upgrade to the AI edition with a higher price tag.

It’s not enough to just be a market leader. Heck, tools like MS Project that have in practice defined the market have been thrown under the bus. It’s all Planner now, with layers of legacy code from different decades running under the shiny new gradient colors. Project Online enterprise customers don’t get the promised parity, instead they’ll get a few years to plan how to migrate away from the MS platform.

Any software ship can be set to fire when you’re a giant the size of Microsoft.

The Copilot Company, managed by Copilot

It’s not like Microsoft would be the only tech giant betting big on AI as the future of everything. As I wrote earlier, the lack of any meaningful profits hasn’t stopped players like AWS, Google, and MS from investing heavily into data center capacity for running GenAI services. When everyone’s playing the same game, how could you step away from the table?

Microsoft said back in January that “our AI business has surpassed an annual revenue run rate of $13 billion, up 175% year-over-year.” So, of course there’s some money flowing in, thanks to all the AI workloads running on Azure. However, keeping in mind that OpenAI is the biggest spender by far, and they’re running a lot of their services on Azure, with credits given to them by Microsoft - well… Is it an infinite loop that just keeps generating more growth for all parties, as long as you don’t pay close attention to who’s paying the bills?

If there was a certain way to make money in the AI business, it wouldn’t be a big bet, now would it? The Redmond crew could always be spending their capital on boring stuff like challenging SAP and Oracle in the ERP market. Where customers are paying real money to build and modernize real systems that are already managing business processes with real tangible impact to the customers’ bottom line. Even though that market also keeps on growing, it wouldn’t be enough of a growth story for the $3.3 trillion software Goliath.

So, the future must be invented, rather than relying on the traditional way of using software in businesses. Satya Nadella is certainly doing his best to live on the edge and use Copilot in a way that consumes his life inside and outside the office, as described by Bloomberg. He sure thinks it’s a cool future to pursue - no matter how sad it might sound to the general population.

The above paragraph from the Bloomberg article was doing the rounds on Bluesky and Mastodon. While it was taken out of the bigger context, the reactions and comments were quite brutal. I don’t think people reacted this way to iPhone early adopter CEO’s that wanted to use their shiny phones for everything. They might have had a “that guy is just a sad geek” reaction to someone walking around with an Apple Vision Pro headset, though.

I think the reactions of regular people matter a great deal. Same way as Satya believes that the tech achievements from five years ago no longer matter. Because it’s never about whether something would be technically possible to build. It comes down to the simple question: do people want this in their lives?

This is different from the question “can GenAI be useful” - to which the answer is obviously “yes”. Despite this latest wave of tech disruption taking the form of pure software on the end user side, rather than being a new gadget that they need to purchase, it makes me think about the metaverse. Or AR, VR, whatever.

People have been trying to enter a virtual, immersive world for a long time. Or to bring a part of the digital world into our physical reality in a more natural way than carrying a smartphone. It could be fun. It could be useful. It could be a big new business for someone. A day may come when it is exactly that.

What if Satya is wrong about the readiness of GenAI and is misreading the room, just like Mark Zuckerberg did when he bet the whole FB farm on metaverse? After burning more than $60 billion since 2020 on their Reality Labs division’s adventures in metaverse. Meta’s CTO has said the metaverse could be a “legendary misadventure” if the sales don’t finally pick up in 2025.

Even if Zuck may wish he didn’t rebrand his company to Meta, he’s still pretty much safe. Being the founder of Facebook, he’s protected by the dual-class share structure that gives him the majority voting control in the company. Besides, with all the attention now on AI, there’s the perfect distraction for the shareholders that can now convince themselves that they’d been betting on AI the whole time.

What if Copilot is a misadventure?

Things are different in Redmond, and Satya isn’t going to have the same luxury as Zuck in terms of job security. Let’s do some speculation on where things might be heading for the Power Platform in particular if it turns out to be a false start for the AI revolution in business apps.

Subscribe to Plus to read the rest.