Earlier this month several news outlets reported that Cisco was discontinuing Intercloud at the end of March 2017 and moving the workloads to yet-to-be-named cloud providers (most likely Amazon Web Services and/or Microsoft Azure).
Is this a shock? It shouldn't be. Cisco has been non-committal regarding any questions about Intercloud for a while now, so it was just a matter of time before it put the final nail in its coffin.
I did like the original vision of Intercloud, which was to create a world of interconnected, federated clouds where workloads could be free to move between them. This might have made sense a few years ago, but the market has changed.
I believe the infrastructure cloud wars are over. Amazon and Microsoft Azure have won. Building a competitive cloud offering is enormously expensive and just not practical. Microsoft had to build out a global, cloud-scale platform for Bing, and Amazon for its e-commerce business. The incremental cost of turning the underlying platform into one that can be used for business cloud services is nominal compared to building it out from scratch. Google should have been a big player as well but it has bumbled its way around the cloud for so long that I think it missed its window.
Also, I'm not sure that customers want to move workloads between different public cloud providers. Maybe years from now they might, but right now the customer focus has been on building hybrid clouds where workloads are moved between a public cloud and a private data center.
For Cisco, clearly the cost and time required to work with service providers and systems integrator partners to build out a network of clouds just wasn't worth it to compete in a market with two massive incumbents. Instead of throwing more resources at Intercloud, Cisco chose to shut it down.
This raises the question, what does the shuttering of Intercloud mean for Cisco's other cloud platforms, in particular HCS and Spark? In my opinion, I don't believe what's happening with Intercloud is related to Cisco's cloud collaboration initiatives, but there are a few things to watch for.
The vision of Intercloud was to connect the dots between the various cloud platforms, and that vision should still hold true except instead of facilitating provider-to-provider connections, it should focus on public to private. Cisco has been aggressive with HCS, and it's well adopted with large service providers. The Spark cloud is relatively new, but anyone who has been following Cisco even moderately closely knows that Spark is the centerpiece of its product strategy now and into the future.
Earlier this year the company announced Cisco Spark Hybrid Services, which allow users to get the capabilities of Cisco Spark but use Unified Communications Manager, Business Edition 6000/7000, or HCS for call control. Cisco has done a nice job with the integration, and it's so tight that the experience is the same regardless of the back-end calling service. Businesses can use the Spark – HCS hybrid service as a stepping stone to a larger cloud strategy.
One area for improvement between Spark and HCS is making it easier for businesses to purchase. Currently, I believe that customers are required to pay for Spark and HCS separately. Ideally, the entire process of paying for Spark and whatever calling service the customer wants should be seamless. As Spark matures and becomes more pervasive, platform and purchasing interoperability with HCS will become critical to Cisco's ability to cross-sell both platforms. The Cisco Spark Flex Plan, announced in November, should eventually address this issue, as Cisco indicated it has HCS on its roadmap for inclusion in the new pricing/subscription model.
Another issue for HCS is finding a way to get smaller service providers to adopt it. Currently HCS is a turnkey solution that includes everything a service provider needs to roll out a UCaaS service -- including hardware, software, virtualization licenses, etc. HCS customers can deploy it in a number of different models including pure hosted, hybrid, or on-premises with remote management. Because of all the components required, the price tag can be quite hefty; even though it's a turnkey solution, it does require some technical prowess to deploy, maintain, and keep current so businesses can have the latest and greatest features from Cisco's HCS partners. Finding a way to bring the cost down would help smaller network operators, system integrators, or VARs adopt HCS.
One thought is Cisco could put up its own multi-tenant HCS cloud and let UCaaS providers resell the cloud service. A good analogy for this is how BroadSoft goes to market with BroadWorks and BroadCloud. The former is hardware and software infrastructure that a service provider can deploy, customize and manage. The latter is a pure cloud service that can be resold via a white label model. One of the advantages of the BroadCloud model is speed at which services can be rolled out. Once it's available in BroadCloud, it can be immediately offered to businesses. In this digital era, where speed to market can mean the difference between becoming a leader or a lagger, the cloud model might prove to be advantageous to deploying infrastructure.
Cisco's cloud strategy has changed significantly over the past several years. Instead of being focused on federating and creating interoperability between clouds, it is now helping customers build and manage hybrid clouds. In the collaboration market, this means continuing to broaden and tighten the integration points between HCS and Spark.
Follow Zeus Kerravala on Twitter and Google+!
@zkerravala
Zeus Kerravala on Google+
