Editor’s Note: The results are in from North Bridge Venture Partners’ 2012 Future of Cloud Computing Survey. In this interview, Michael Skok, general partner at North Bridge, and Marten Mickos, CEO of Eucalyptus Systems, share insights around the survey findings and also discuss upcoming changes in the platform-as-a-service segment.
SandHill.com: The survey found that 50 percent of the 785 respondents have complete confidence that the cloud is viable for mission-critical business applications. But only 37 percent of that group is customers and the rest are vendors. What needs to happen to increase the customers’ confidence?
Michael Skok: The gap between vendor and user confidence around cloud is based on five key things that vendors still need to address. First they need to do a better job of providing more complete value propositions. They also need to clearly state the business benefits. Third, they need to do more case studies and publish success stories. Fourth, they need to provide more proof of the benefits. Sharing the facts about return on investment is the fifth point.
SandHill.com: Only three percent of the respondents this year stated they believe the cloud is still too risky, compared to 10 percent in the 2011 survey. And 12 percent stated that cloud solutions still need to mature in the security and compliance aspects. Are these large enterprises, small and midsized businesses (SMBs), or both?
Michael Skok: In last year’s survey, responses indicated mostly SMB adoption because of the obvious benefits in economies of scale. This year we’re seeing that, as the confidence in the cloud grows, there’s a shift toward large enterprise adoption. The first driver of that is SaaS because these applications solve business problems (HR as a service or CRM as a service, for example).
Of the 55 percent of CIO respondents who stated their companies increased spending on SaaS in 2012, two-thirds are from major enterprises (10,000 or more employees) and include the Fortune 1000.
SandHill.com: I noticed that the survey respondents listed pricing as an inhibitor to cloud adoption. Isn’t pricing one of the benefits and reasons for cloud adoption?
Michael Skok: Absolutely. What’s happening here is that buyers are challenged with the complexities of pricing when vendors present their cloud offerings. As an example, some cloud vendors charge for the cost of uploading and downloading data; others only charge for storing data. So buyers need to be able to do an apples-to-apples comparison.
It’s also very difficult for them to compare cloud pricing to their internal IT costs because most internal IT is not priced as a service. These complexities make it difficult to project what a cloud solution will cost.
SandHill.com: I noted that the survey respondents ranked vendor lock-in and interoperability pretty high. I thought this aspect of the cloud was improving. What’s going on in this area?
Michael Skok: According to the survey, that’s not the case. Vendor lock-in and interoperability are actually the same issue in many ways. The issue is that when people see things like a cloud outage at Amazon as happened last week, the question they ask themselves is: How can we spread the risk and not put all our eggs in one basket? They can only do that if they have the advantage of being able to work with multiple vendors. That requires interoperability and not getting locked into one stack. I think this issue will continue for quite a while.
Michael Skok: The challenge here is that right now we’re still early in cloud, so people are innovating as quickly as they can to meet customer needs. Because they’re so busy innovating, the last thing they have time to do is look over their shoulder and figure out what their competitors are doing and how to interoperate with them.
SandHill.com: When will this challenge change? Three years?
Michael Skok: It will change once we reach what I would describe as the tipping point between innovation and adoption. And that is likely to be some number of years ahead.
SandHill.com: More than five years?
Michael Skok: These kinds of major industry shifts take several years. Here’s the trade-off, in my opinion, which is informed by both the survey and conversations with CIOs in customers and vendors on an ongoing basis. A year ago, vendor lock-in and interoperability were important, but it’s now becoming more important as the survey highlighted. The more important it becomes, the more people will focus on it with vendors. That’s why I believe the open source stack is getting so much attention. I think any open source play will solve it sooner, maybe in two to three years.
With the major vendors it may be three to five or more years. The reason is that the vendors are so focused on innovating as mentioned earlier. To some extent, the more that they build, the more they define the standards. For example, Amazon now has a de facto set of standards based on their market dominance; and the more they get adopted, the more they become the de facto standard. In many ways, there’s not much incentive for Amazon to do anything other than innovate to stay ahead of the pack and let the standards get defined as de facto around them.
But for other vendors, it’s going to become increasingly important for them to figure out how to interoperate. So that’s why companies like Eucalyptus are emerging and why players like Rackspace are driving for OpenStack. I think the vendor powerbase will change dramatically over the next year.
SandHill.com: The survey found that hybrid cloud adoption will almost double over the next five years and that both the public and private cloud models will shrink in adoption. Why isn’t there more adoption of the hybrid cloud model now?
Michael Skok: A lot of the original investments in cloud solutions were around virtualization to get the cost savings out of servers. People take the step from there to the private clouds. But the step from private cloud to public cloud in order to get the advantages from economies of scale is a big step because it means giving up control. That challenge is still shown as a big inhibitor in the survey. The result is people take a half step and go to hybrid cloud.
Marten Mickos: Overall, I think nothing will shrink. All cloud segments are growing at an astonishing rate. But the proportions may change. We are seeing customers going forcefully for private cloud solutions, but only if they know that there is a hybrid option for the future. We have customers running hybrid clouds today, but we have even more for whom it is an important goal for the coming years.
When large corporations take the step from a virtualized environment to a full cloud solution, there is a need to train staff and redesign some processes. That’s why the adoption is not as fast as one would think when looking at the offerings available. First you establish a private cloud, and then you start experimenting with hybrid scenarios (of which there are several).
SandHill.com: I think one of the most interesting survey findings was the high projected growth of both the Platform-as-a-Service (PaaS) and Infrastructure-as-a-Service (IaaS) models over the next five years. What is driving that growth?
Marten Mickos: We see strong growth in IaaS and we think the reason is that the model has been so thoroughly validated by AWS. A huge ecosystem of software, developers and companies has emerged around AWS and their APIs (EC2, S3, EBS, etc.) These practices spill over into enterprises and allow them to see the benefits of going cloud — private cloud, hybrid cloud or public cloud. It is not unlike the shift we saw about 10 years ago from old-fashioned solutions to Linux-based environments. One could even argue that AWS is the Linux of today.
Michael Skok: People tell us the driver at the PaaS level is about reducing complexity. Picking a platform, selecting the components and developer tools for it and configuring all of those, optimizing them, and then deploying it and maintaining and updating it is a lot of complexity. PaaS takes all that away and reduces it to the simplicity of online development and deployment.
There is a shift happening at the PaaS level that is really interesting. Amazon is a good example. They started as a public IaaS player and have continued to move more and more up to PaaS because they’re adding capabilities. We talked to them, Rackspace and players about this. They all recognize that in order to differentiate themselves from other IaaS players, they need to add more capabilities.
What people told us is that at an infrastructure level, things like storage (for disaster recovery, backup, business continuity) are big drivers. If you’re not accessing data in real time, on premise storage can be expensive. The cloud allows for much cheaper storage. So these kinds of services will continue to drive IaaS.
But here’s the rub. I think we’ll see all the infrastructure players becoming more like platform players. We’ll see a squeeze: the IaaS players at the bottom will move up, and the SaaS players at the top will move down. It’s not clear to me that the PaaS category, long term, will end up as a distinct category.
SandHill.com: So the PaaS model may not survive; it will just become part of IaaS at the bottom or part of SaaS at the top?
Michael Skok: Obviously PaaS is not going away. But, yes, I’m making a bold prediction that PaaS will be absorbed from either IaaS at the bottom or SaaS at the top.
Think about Salesforce. They’ve already spent hundreds of millions of dollars on building a platform to deliver CRM as a service. That platform has many components in it that would be usable as a PaaS. A good example is Force.com and then how Salesforce is offering database as a service as a PaaS offering. Amazon has added similar capabilities onto their IaaS with DynamoDB. I think we’ll see more companies taking those capabilities on which they built applications and enabling them to be accessed as a platform.
Slide 30 in the deck on the results of the survey shows very clearly how Salesforce has moved down and Amazon has moved up. And VMware is trying to move in. Players like Oracle just entered the space. Microsoft with Azure is very much in the space with a huge .Net legacy developer base. It’s going to get very competitive. And it will be tough for new entrants coming into the space to carve out a niche with just a PaaS.
Click here to view a slide presentation of the results of North Bridge Venture Partners 2012 Future of Cloud Computing survey. SandHill is one of 39 survey collaborators.
Michael Skok joined North Bridge Venture Partners in 2002. Michael has been a software entrepreneur and CEO for 21 years. He founded, led and attracted over $100M in private equity to investments in several successful software companies, and built a significant multiple of that in value for investors as a VC. Michael focuses on market-changing technologies such as cloud computing and disruptive business models such as open source. He created the first industry-wide open source survey, now in it’s 6th year; in SaaS he led CIO forums as early as 2003, culminating in his work at www.futurecloudcomputing.net with the first multi-vendor industry wide survey and collaboration on the Future of Cloud Computing. Michael also focuses on applications in ecommerce, enterprise mobililty and Big Data. Representative investments include Acquia, Akiban, Apperian, Demandware, Unidesk, Actifio and Revolution Analytics. Contact him at email@example.com and www.mjskok.com or on Twitter @mjskok.
Marten Mickos is CEO of Eucalyptus Systems, providing open source cloud computing platforms for on-premise use. Previously as CEO of MySQL AB, Marten grew that company from a garage startup to the second largest open source company in the world. After the acquisition by Sun Microsystems of MySQL AB, he served as senior vice president of Sun’s Database Group. He is a recipient of the Audemars Piguet “Changing Times Award: European Entrepreneur of the Year 2006.”
Kathleen Goolsby is managing editor at SandHill.com.