Thursday, June 27th, 2013

The Rise of the Cloud Stack

By

Something changed this week in the enterprise software world. In an industry known for ruthless competition, a number of players – Microsoft, Oracle, Salesforce and NetSuite – introduced partnerships that portend a very different future.

In very un-Larry Ellison-like fashion, something akin to harmony was proposed, “when customers choose cloud applications, they expect rapid low-cost implementations; they also expect application integrations to work right out of the box–even when the applications are from different vendors.”

This is a striking development in the enterprise software world. It also happens to be a positive one – for Oracle’s future relevance, for independent cloud vendors, and most of all, for customers. For the last decade, the enterprise IT landscape has been a battleground of competing paradigms. Oracle has pushed a vast empire of all-or-nothing, fully integrated solutions, while comparatively fledgling cloud vendors have offered choice and integrations with other best-of-breed services. And customers, to a large degree, have been forced to choose between these divergent philosophies and realities. In one swift move, the enterprise software world has gone from being Game of Thrones to Friends (or, at a minimum, Frenemies).

friends

Perhaps most importantly, Oracle’s moves this week are not just a blip or even that anomalous; they follow a well-established pattern in the technology ecosystem, of oscillation between interdependent, integrated solutions from a single vendor and openness from a larger set of players. And with the poster company of consolidation this week embracing a more open world, the industry’s progress towards a new model in the cloud era is accelerated.

Fully Integrated vs. Best-of-Breed

To understand what is happening now, we need to look back at the history of the technology landscape. While no period is ever fully defined by one overriding approach (e.g. in mobile phones we have Apple, the quintessential vertically-integrated company and Google, the ‘open’ disruptor, both simultaneously experiencing some form of success with their differing approaches), ecosystems tend to follow a certain pattern.

This dates back to the very beginning of the computer industry. Andy Grove explains that companies in the mainframe and minicomputer era “competed in this industry as one vertical proprietary block against all other computer companies’ vertical proprietary blocks. Salesmen would show up and offer their vertical combination of things, and the company they were selling to would decide to buy one proprietary line and not the others.”

There was a very compelling reason for this. Because of the level of integration required between components that made up mainframe computing and the lack of industry standards in place, vendors had to source, build, and define most of the components (software, hardware) themselves to effectively compete. The key to the vertical stack was about reducing the complexity in the customer proposition of combining disparate systems, thus driving more sales, faster. However, at some point, as technology becomes good enough, we tend to see ecosystems move toward open and ‘modular’ approaches.

With the rise of Win-Tel, instead of buying an entire vertical stack of hardware, operating system, and apps from a single vendor, we moved to a world where you accompanied best-of-breed chips with a packaged computer and chose a variety of software solutions to run on top. Where there were few network effects or an abundance of standards, we saw vibrant competition; namely in the software application space. This led to rapid innovation and a constantly churning computer industry, with leadership positions changing hands every few years among software companies like Siebel, PeopleSoft, BEA, Great Plains and Lotus.

Rise of the Monolith

Eventually, however, the industry reached a breaking point: the rise of heterogeneous software in the client-server era meant that for every new technology implemented, the CIOs and IT teams had to become experts in an all new set of infrastructure, services, and integration. This made supporting and deploying solutions from disparate vendors cost and time-prohibitive.

Larry Ellison recognized this as an opportunity to get customers to buy all their technology from a consolidated provider, noting the horizontally-integrated approach simply wasn’t working: “It’s incredibly difficult and expensive to make these systems communicate at all – and it’s impossible to make them communicate well.” And with this rejection of integrating disparate systems came an era of mass consolidation in the enterprise software industry. In the interest of market share, dozens of the best-of-breed leaders were gobbled up.

In some respects, this trend ostensibly served to benefit customers: the same ‘solutions’, fewer vendors to call, and the promise of a fully integrated suite of products where everything magically worked together. And it was a great period for the giants – Oracle, SAP, Microsoft – as they captured more profits and control of the ecosystem. But ultimately, it did little for enterprises and the products they were using. With limited competition comes limited innovation.

The cloud, over a decade later, is finally threatening the approach that was so pervasive as the primary growth mechanism in the early 2000s.

The New Cloud Stack

In the on-premise era, integrating enterprise solutions entailed vast amounts of complexity for every new solution implemented. Getting your CRM to work with your ERP or HRM from varied vendors was a near-impossible task and led to overdrawn budgets, delayed projects, and incomplete systems.

But with the cloud, a customer can light up any application near instantly – a user of Workday no longer has to worry about the servers, storage, networking or middleware that goes into delivering a world-class HCM system. It all comes for ‘free’. And given the low friction, thousands of enterprises are now running their businesses through a mix of best-of-breed technologies: HCM on Workday, CRM on Salesforce, ERP on Netsuite, Support on Zendesk, Social on Jive, Business Intelligence on GoodData, and so on. Instead of relying on a single vendor to do all their jobs, customers get the best products in each category to solve their problems.

CIOs who’ve embraced this approach simply don’t need one-stop shops for all their technology needs anymore. The cost and time burden to implement new solutions and integrate them becomes incremental (as opposed to exponential), and with emerging identity and integration solutions like Okta, Ping, and SnapLogic the burden is reduced even further.

And integration in the cloud era has become as simple as an API connection. Most cloud services have created a web of interlocking parts: an update in Salesforce prompts an action in Zendesk; a new record in SAP shoots off an update to Yammer; a document in Box can be accessible from Jive. This is the direction the world is moving. From this diversity of systems, customers will get choice, and with this choice we’ll see better applications and solutions emerge. At a lower cost, and at a higher quality.

Previous to this week, the industry was faced with two competing paradigms: a war between the ‘fully integrated’ suites and the best-of-breed cloud stacks. Customers were left to make a philosophical decision, risking going down a path unsupported by some portion of vendors. With this week’s encouraging moves from both Oracle – the leader of monolith era – and Microsoft, a single paradigm of open solutions may fully win out. It’s an acknowledgement that enterprise software is no longer zero-sum; with expanding markets and an internet population of billions, there’s more opportunity than ever before.

A new cloud stack is rising. And this is good for everyone.

By

See all of Aaron's articles.

  • Anand Gautam

    Great article!