As the enterprise software market slowly morphs into the enterprise software and platform market, it’s become necessary to carefully define what it means to be a successful platform vendor. Importantly, that definition has nothing to do with technology – okay, maybe a little – but it does have a whole lot to do with people and perceptions.
A platform is really the foundation of an ecosystem, and in the platform wars, it’s the best ecosystem that wins. Nothing else matters. So, when looking at the three top contenders for the leading enterprise software platform ecosystem – Microsoft, SAP, and Salesforce.com – it’s important to make sure we understand what having a successful platform and ecosystem really means.
My sense of what makes platforms and ecosystems successful places these companies in what today is a three-way race to the top. Other top contenders – such as Amazon and Google – just don’t have the enterprise DNA to become major players: their consumer tech chops have led them to offer great infrastructure services that have them firmly engaged in a feature/function/price race to the bottom. A latecomer to cloud platforms, Oracle isn’t a serious player – they’re about to muck up their already mucked up platform plans with the acquisition of a multi-platform mess called Netsuite. IBM’s inane focus on Watson and its focus on everything being a consulting gig – no packaged software, please – just won’t make it in the long run either. And Infor, which has the right strategy and technology for the job, so far lacks the customer proof-points that would bring its ION platform and Cloudsuite offerings into the top tier.
But Microsoft, SAP and Salesforce all have some serious assets, and serious aspirations, to be highly successful players in the platform-cum-ecosystem world. And all three have serious – and very different – challenges as they march towards what today is still not a winner-take-all, zero-sum race to the top. So far.
Defining what constitutes a successful ecosystem is simple, using my lying awake at 2 am in the morning definition. When software developers, entrepreneurs, or VCs wake up at 2 am with a hot new idea for a killer app or product, the platform they reflexively chose for building their app is either the easiest platform for them to get started on or the one that provides them access to the very best ecosystem, preferably both. When thousands of developers, entrepreneurs, or VCs make the move to the same platform, a successful ecosystem is born. Move millions, and you’re the platform/ecosystem leader.
What’s behind these reflexive motions towards one platform or another? At this point in the evolution of the market, a successful platform, theoretically, fulfils the full lifecycle of the successful product or service, from nascent ideation to order fulfillment and payment, to renewal and eventual end-of-life, rinse and repeat. This includes a list of functions and capabilities that starts down at the technology level with developer tools, databases, access to services, middleware, and APIs, and finishes at the end-customer level, when a product is purchased and paid for, and implemented or configured in as much of a self-service model as possible.
In between are a host of important functions: open APIs, low-cost or free dev and test environments, a wide choice of dev tools, and databases, and the elasticity to start small and scale big. Cost transparency is also very important – who wants to build a killer app that’s too expense to make a killing on – and a super easy on-ramp for both developers and eventual customers is also a requirement. I’ve seen some platform pricing schemes that look like the spawn of Rube Goldberg and the Code of Federal Regulations: complex, overwrought, guaranteed to send a developer running to mother Amazon in tears. Not a way to succeed in the platform wards.
Increasingly, mother Amazon notwithstanding, having an extremely elastic cloud is just table stakes. What matters more than raw infrastructure is access to specialized business services: APIs for business services that support standard backoffice functionality as well as supporting nascent areas like IoT and machine learning. APIs that connect cloud assets to on-premise assets. APIs that support complex, hybrid environments with mixed legacy and leading edge functionality. APIs for specific industries, geographies, and regulatory requirements. And on and on. The bigger the shopping list for extending applications to the rest of the business world – that vastly heterogenous, mixed up, often functionality irrational, real world that enterprise software customers actually live in – the better.
Then there’s the customer side. There needs to be a massive customer base eager to solve their digital transformation problems at the same time they engineer an upgrade from the primordial ooze of Excel and paper processes. Those customers need to be given a killer user experience that hides the underlying chaos that leading platforms are trying to control. And they need these experiences to be self-service, intuitive, intelligent, and as prescriptive as possible.
Perhaps most importantly, there needs to be a buzz, a coolness factor, an ineffable momentum that draws developers, partners, and customers into the platform’s magical orbit. In a funny way, platforms and ecosystems today need to think more like consumer brands than technology brands, with a focus on the millennial demographic, an ability to exude excitement and vigor, youth and hipness. An ability to be cool, quick, nimble, ever-evolving. A hack on the staid, old enterprise software model of yesteryear.
Finally, there has to be a serious investment on the part of the platform/ecosystem company on making the above happen, particularly the buzz/coolness/magic part. That involves money, it involves senior executive spokespeople rallying the troops, it involves creating an almost religious revival movement that looks antithetical – but isn’t – to the dominant nerdy, basement-dwelling engineering culture that enterprise software was born to.
So, armed with these basic criteria, here’s my assessment of where my leading platform/ecosystem vendors stand. Right now, I’d say it’s a pretty close race, with each one running ahead in certain criteria and behind in others.
Microsoft’s investment in Azure, and the resounding success of Office 365, has placed the company’s platform and ecosystem in serious contention. They have an amazing array of services – database, IoT, machine learning, and the like, some new and very appealing hardware assets like Holo Lens and Surface Hub and a new generation of Apple-butt-kicking PCs, all drawing on the universality of the Office user experience when and where it makes sense. Their recent Dynamics 365 announcement was further proof of where they are heading, which is in the right direction, though the mid-market origins of Dynamics AX means that they’ll have to work hard to legitimize their ecosystem as capable of serving the global, large enterprise market.
Nonetheless, Azure’s Lifecyle Services is without a doubt the best cloud ALM tool in the market, and as the complexity of hybrid cloud environments grows, LCS is going to be an increasingly powerful asset and competitive differentiator. Microsoft can clearly grow into the top tier of the market, but it will take some time and effort.
The big issue for Microsoft, ironically, is that the company is still trying to pivot its ecosystem in the right direction. While there are millions of developers who claim .NET and other Microsoft credentials, the enterprise software ecosystem of the future isn’t necessarily the place where the classic Microsoft developer – tech-heavy, but business light – can excel (spot the bad pun, win a prize J). Microsoft’s ecosystem still suffers from an on-premise, PC-based bias, which means it needs a new class of developer to step up to solidify that top tier position. There’s every reason to believe they’ll make it – Lord knows they know how to build an ecosystem, and Microsoft isn’t shy about spending real money to maintain it – but a developer shift of this magnitude will also take time and effort.
SAP has the top spot when it comes to large enterprises and highly complex environments, though SAP also lacks the ALM tools to support this position. But the capabilities of the company’s HANA Cloud Platform, especially its growing API library, IoT and machine learning capabilities, means that SAP is well-positioned to succeed. S/4HANA, on-premise and in the cloud, when combined with SAP’s myriad cloud assets – Ariba, SuccessFactors, Concur, Fieldglass, and hybris, among others – makes for a compelling platform on which to stage complex business processes that span multiple lines of business and products. In particular, as more and more customers move to a relatively restrictive, one-size-fits-all cloud world, HCP and its tools and services can become the place where customer/industry/geography-specific functionality can be built and staged without breaking the “no blueprint” cloud model.
But SAP also suffers from a paucity of the right kind of developers, and, more importantly, a limited quantity of internal DNA that understands what it takes to reach this new generation of developers. I was recently at SAP’s TechEd conference in Barcelona, and it was a classic case of a vendor pitching tomorrow’s message to last year’s audience – a problem endemic to all vendors transitioning a new platform/ecosystem paradigm.
I think the message that Bernd Leukert, the member of the SAP board in charge of products and innovation, delivered at TechEd was spot-on relative to the new platform/ecosystem challenge, and for the most part I think the Tech Ed audience got it. But the audience, by and large, was old guard SAP partner/developers. The new guard, which does show up for SAP’s Innojams and other developer mini-events, was definitely at Tech Ed, but they were in the minority.
Ultimately, SAP needs to do more, much more, to reach out to a new audience and make that reflexive move to SAP and HCP a given, not an afterthought. SAP’s recent agreement with Apple to unleash the IoS developer ecosystem on SAP’s platform is start, but IoS developers are like traditional .NET developers when it comes to next generation enterprise software: only a few can really step up to the new platform challenge.
Overall, the problem of building a name for itself outside the SAP ecosystem, particularly in Silicon Valley and other tech-centric locales – combined with SAP’s historic allergy to spending the money it takes to build a real developer ecosystem – will be the main gating factor for SAP’s platform and ecosystem aspirations. The spirit is willing, but it’ll take more than good intentions and the best business APIs in the business to make SAP’s ecosystem number one.
Last but not least is Salesforce. Unlike SAP and Microsoft, Salesforce has cracked the developer ecosystem code with Trailhead. While the company’s recent Dreamforce conference was, vis-à-vis the company’s ecosystem aspirations, another example of the “next year’s message, last year’s audience” problem, the Trailhead zone at Dreamforce this year was the exact opposite, a shining example of how to get the platform/ecosystem strategy right: stuffed with people, buzzing, standing room only, gamification to keep things rolling, and lots of woodsy icons to keep the buzz from ever getting too serious. And, most important, present and accounted for was a real ecosystem excited about using Salesforce’s platform(s) to drive real change in the enterprise.
The irony of this platform buzz leadership is the relative paucity of high level business services that Salesforce.com has to offer. The company’s vow to be the very best CRM platform sells its post-CRM developers and partners short: while one could argue that the customer is and will forever be the center of the business universe, both SAP and Microsoft offer a broader vision of the enterprise that is likely to appeal much more to those 2 am in the morning insomniacs who are looking to develop the next killer enterprise app.
If all you want is to build off the existing Salesforce.com base, which is considerable and intersects heavily with both SAP and Oracle, then the Trailhead ecosystem is where you want to be. But if you need more – a broad range of business services and APIs outside of a core CRM system, as well as comprehensive ALM tools and the like – then Salesforce.com might not be the best place for your efforts. Trailhead will support robust functional designs and can appeal to developers and citizen-developers like no other platform, but right now the range of capability is limited. For now. Salesforce.com’s platform aspirations are clearly headed in the right direction, and its Trailhead developers are engaged and ready to get going: that’s a pretty potent combination, and one that will keep Salesforce in the running for some time.
The bottom line is that these three top tier vendors are equally strong for different reasons, and that, due to the nascent nature of the platform/ecosystem opportunity, that’s just fine. A three-horse race will make every vendor run faster and try harder. Which means that the ultimate winner, the one that really matters – the customer – will reap the benefit of a classic competitive market evolving into a nascent and highly important new form.
Looking at the sources of revenues for all three top contenders, the transition to platform is still to come. Which is also good for all concerned: Customers aren’t actually as ready as their vendors would like to think, and the vendors still have a lot of kinks to work in their strategies and offering. Time is on all of our sides right now – the quarterly cadence demanded of the public equity markets is going to have to be patient.
But the platform/ecosystem game is clearly afoot, and the transition is as inevitable as previous transitions – to client/server computing, Internet commerce, and mobile, for example – have proven to be. Which further argues for a good dose of patience. Even in the compressed timeframe that technology markets now lives in, the shift will take time for all – vendor and customer – to absorb.
The tech side is easy, but when it comes to people and perceptions….that stuff just takes time.