At a press/analyst meeting last spring, Salesforce.com CEO Marc Benioff was asked whether he had any plans to build out an Amazon AWS-like capability to complement the rest of his cloud strategy. His scoffing reply was right on the money. Competing with AWS and other commodity-level cloud services was “a race to the bottom,” Benioff replied. Case closed.
The cloud market has another race in the works, however, one that Salesforce, SAP, Microsoft, Infor, and others are all contending in with increasingly ambitious strategies, new services, and strategic acquisitions. It’s the race to the top of the cloud, up in the high-value business services layer where the unique capabilities of the cloud, aided and abetted by the commoditization of IaaS, PaaS, and other low-level services, can enable a set of increasingly high-value services and capabilities that render the cloud the place to be for business innovation.
This race to the top is taking two highly complementary forms. Form number one is the race to build out a business network strategy that promises to support the complex, matrixed networks of business interactions that have outstripped the inside-the-firewall, linear business processes – and software – of the 20th century. Form number two is the race to build out a set of componentized business services, cloud natives each and every one, that can be compiled into end to end services, or attached as services to the aforementioned business networks, or used to extend existing on-premise applications into the cloud. Both forms are hugely important in the quest to deliver genuine, and highly valuable, innovation in the cloud, and both are harbingers of a major inflection point in the value of the cloud in the enterprise.
SAP was the latest batter to take a swing at the race to the top. Implicit in its recent HANA Vora announcement (a name that sounds like the perfect nom-de-guerre for a beret and sequins resistance fighter for truth and justice, though I’m guessing it’s based on the latin root vorare¸ which is also the root of the word devour) was the ability of SAP to unleash a set of business services, developed in house and via partners, that will be used as building blocks for the next generation of cloud business apps. The four that SAP announced, marketing, commerce, and loyalty management services, all based on hybris, and a separate global/local tax calculation service, are meant to be the tip of an iceberg of easy to deploy and easy to consume cloud-based business services.
HANA Vora also ups the ante for developers, adding to the nerd’s dream full of dev tools that are piling up in HANA by adding native access to Hadoop and other tech wonders. This is a truly awesome milestone for HANA, don’t get my somewhat flippant attitude wrong. But tech is as tech does, and the real goal of the HANA Vora announcement is to use the HANA Cloud Platform to relaunch the composite apps market SAP tried to create back in the early days of the 20th century with NetWeaver and xApps, with hopefully much better results.
The xApps concept never took off, mostly because the complexity of supporting the necessary integration and orchestration on-premise was too onerous for the service-oriented architectures that were the basis of these efforts. Perhaps more importantly, xApps and composite apps also violated my first rule of enterprise software – the hardest thing to do in enterprise software is to ask someone to do something they’ve never done before. In this case the violation was primarily in regard to a developer community that had trouble imagining just what a set of killer composite apps would look like and do, and secondarily in regard to a user community that at the turn of the century never really understood the appeal of composite apps in the first place. IT got it, but the LOB wasn’t really on board.
Putting the composite apps concept in the cloud, by making these composite pieces universally accessible via a cloud-based set of APIs, brings composite apps back to the realm of not just possibility, but reality. At least in theory. The violation of my first rule of enterprise software still stands, as does the inherent violation of my second rule – the easiest application to sell is the sleeping pill that lets the user who’s losing sleep over a specific business problem get a good night’s rest. The problem with the second rule is that if the sleepless don’t know that there is already a sleeping pill, in this case composite apps, that will help them sleep, they can’t know to ask for them. So far, that cognitive connection has yet to be established in the market, and this makes establishing this connection, composite app as LOB sleeping pill, particularly among developers, SAP’s and its erstwhile competitors’ biggest challenge.
Of course, SAP isn’t the only vendor in this version of the race to the top. Microsoft is there as well, although it too suffers from the same sleeping pill problem. Microsoft has been pushing a broad set of componentized, cloud-based business services (albeit much too quietly and surreptitiously) via its Azure Data Market for some time. The 300-plus services in the Azure marketplace (things like HealthMethods Provider Metrics and Customer Churn Prediction) are testimony to the interest in the concept by the providers of these services, but it’s pretty clear that the developer community that in theory would be all hot and bothered by this collection is largely missing in action. At least so far.
Microsoft’s higher end cloud business services also include putting the broader Microsoft desktop and server portfolio – products like Office, SQL Server, Sharepoint, and Power BI, among others – into Azure for use as services accessible by cloud-based solutions. Dynamics’ AX enterprise software suite and its Lifecycle Services will also serve as building blocks or services destinations for new cloud-based “race to the top” applications and services. The fact that Microsoft will soon announce AX 7, which will run in Azure (as well as in private clouds and on-premise) adds some serious enterprise software cred to Microsoft’s odds.
So far Microsoft is contending only for the business services form of the race to the top. Microsoft’s plans to support complex business networks, assuming there’s some thinking going on, remain unannounced.
Infor is in the race to the top as well, squarely focused on the business network version. Its recent acquisition of GT Nexus is effectively a $675 million bet that, as CEO Charles Phillips told the analyst community when the deal was announced, “The future will belong to networked companies.”
The GT Nexus deal is an interesting one. GTN helps manage global logistics supply chains in the cloud, enabling companies to track and trace the movement of supplies from loading dock to shipping container to store shelf. They do a pretty good job of it, better than most, and are able to provide a degree of visibility, by tracking the different documents – bills of lading, customs documents, POs, etc. – that is hard to do even with a RFID tag stuck on every shipment.
More importantly, this visibility provides GTN, and now Infor, with a gold mine of data about the performance of the different players in global supply chain. Analyzing those data will allow Infor to continue GTN’s move into an important business network service that its rivals aren’t really doing as yet: trade finance. If a manufacturer can prove its bona fides as a top notch supplier trading through a highly efficient logistics network, it can demand, and receive, significantly better credit terms on the loans it needs to build and ship the products it is contracted to produce. This brings much needed capital at better rates to business network members, and should in many cases more than justify any investment in time and money that Infor customers make in signing on to the GTN story.
These kind of metadata analyses, which can only be done at scale in a business network, are a main reason why there’s so much interest in the network version of race to the top. Infor’s Phillips, whose ties to the banking and finance world are local (his New York office is a couple miles from Wall Street), knows that the current global trade financing crisis will drive a lot of net new business and net new partners to Infor and its fledgling business network.
GTN fits nicely into Infor’s Cloudsuite strategy, which is focused on providing over a dozen cloud-based suites of services geared to core industries, such as automotive, aerospace and defense, hospitality, fashion (which dovetails with the large number of fashion and retail customers using GTN), healthcare and others. The organization of cloud functionality into industry-specific offerings is a smart strategy for Infor, which needs to provide an upgrade and innovation path that is as short and sweet as possible so that the relatively few pioneers in its otherwise laggard-heavy customer base can quickly demonstrate the wisdom of moving off a legacy Baan or Lawson implementation into a sleek, innovative, rapid-time-to-value cloud solution. GTN will be a nice enticement for a broad swath of Infor’s customers that are going to have to make move, and do it soon.
While Infor is definitely contending in the business network side of the race, SAP is currently in the lead, with its Ariba Business Network strategy the most fully formed of the lot in terms of the number of companies in the network, 1.7 million, according to SAP (versus 25,000 in GTN, though GTN’s services are much more strategic than the indirect procurement that Ariba specializes in). Ariba helped pioneer the business network concept over a decade ago (around the time SAP was testing out composite apps for the first time) with its indirect procurement offering, and is now moving towards the more interesting, strategic, and complex direct materials market. When you add Fieldglass’ contingent labor procurement function, Concur’s travel and expense function, and elements of SAP’s other cloud properties – S/4 HANA, SuccessFactors, among others announced and as-yet unannounced – there’s certainly a lot of potential network functionality in the portfolio.
But making the whole of these properties greater than the sum is still very much a work in progress. They all need to migrate to the HANA Cloud Platform and then go through the even more complicated process of orchestrating what are today cloud silos of functionality. SAP has been diligently moving in this direction all year, and is definitely focusing on cross-pollinating its cloud offerings as much as possible. This always looks easier than it is, but the political will is definitely there to break down the silos at SAP, despite the somewhat confusing fact that Ariba, Fieldglass, and Concur live in one part of SAP’s cloud hierarchy, while SuccessFactors and other cloud properties live elsewhere.
Regardless, the business services side of the HANA Vora announcement can also play a role in extending SAP’s Business Network ambitions. If you’re building out a fully formed procurement network, you need things like tax calculation services, among many others. (You might also like some version of Postcode, for example, from Azure’s much more abundant service offerings. Maybe these two platforms should figure out how to interoperate and leverage each other. Just sayin’ a little coopetition might be a good idea….) The dev tools side of HANA Vora also help, and together they give SAP the means to extend its Business Networks by leveraging its considerable ecosystem, something every vendor will need to do – even if there are still some great properties like GTN to be acquired.
Ultimately, the vendor momentum is exciting and important, but the problem with both ends of this race to top is the problem of market maturity. The business services side desperately needs a horde of developers to start coding their butts off building new cloud services based on a composite model. With HANA Vora, Microsoft Azure, and the Amazon AWS services that underlie Infor’s Cloudsuite, the vendors are setting a pretty rich table for the developer community.
But what’s needed now are legions of composite app and business network savvy developers, and I’m not sure a critical mass of these developers exist today. I saw this lacuna at Microsoft’s Build developer’s event last April: there was no shortage of great programmers, but I came away from Build with the feeling that there was definitely some market maturity needed on the developer side. I’ve felt this in conversations with top notch NoSQL developers, who consider themselves, quite rightly, at the bleeding edge of today’s tech world but are all too often focused on coding new cloud services instead of looking for and using third party services. And I’ve seen it in the disconnect between every vendor’s IoT dreams and the ability of the developer side to imagine, much less build, the killer IoT apps that, by their very nature, must span multiple, cloud-based business services and appeal to industry-specific business networks and their industrial customers.
So it’s important to make sure the expectations for this race are well set: This isn’t like the early days of the mobile app market, when every Dick and Jane programmer was off building little mini-apps for iOS and wowing the masses with light-weight, one-trick-pony apps. The componentized services that race to the top developers should be building and using may be mini, but the apps that are expected to be built from them are decidedly maxi. And that takes a kind of developer that is in short supply today.
Nonetheless, the vendors have to start somewhere, and it’s safe to say that Microsoft, with Azure and its panoply of services, is the current leader in the services version of the race to the top. And SAP, as I stated previously, is leading on the business network side. But Infor is there too, a vendor that too many competitors chose to ignore, to their peril. And don’t forget Salesforce, which is playing hard for market share in the services side, though in my opinion there are too many core capabilities that have to come from the ecosystem – as opposed to organically within Salesforce – for their services play to compete well in the market.
Regardless, it’s hard to handicap a race when the competitors have just left the gate to run a race that’s never been run before. And, to torture the metaphor a little more, considering it’s a race with no discernable finish line, I wouldn’t honestly bet on any single vendor to win, place or show as yet.
There will be one solid set of winners, and that will be these vendors’ customers. The race to the bottom is lowering IT and infrastructure costs as it is simultaneously improving the interoperability in the cloud that is necessary to move from the current silos of cloud functionality to a world where business processes span cloud silos as easily as they span internal silos (it’s easier, actually, the internal silo problem is still a major problem.) This PaaS and IaaS race to bottom will make the business services in the race to top easier to consume, and more importantly, safer to consume, as the ability of vendors to lock in customers will diminish as their cloud platforms’ interoperability improves.
What’s most important is that the true value of the cloud is emerging, what I long ago called SaaS 2.0. The cloud’s true value exists in leveraging its potential as the place to deploy the applications and services that define the brave new world of interconnected, global businesses and networks, rather than just flipping on-premise apps into the cloud for easy deployment and out of a cap-ex financial model. The moves by these and other vendors to provide the development and deployment environments that can enable these net new processes and business networks are right on the money: the cloud will revolutionize global business in ways we’re only just beginning to imagine. And it’s all starting now.