Up in the Cloud: Hype and High Expectations for Cloud ComputingPublished: January 16, 2013 in Knowledge@Wharton
Cloud computing is a much hyped but often misunderstood technology that is gaining traction in different industries around the world. Businesses are integrating the cloud into countless systems, from HR to finance. Full adoption and acceptance of cloud computing, however, are still far away.
A recent global survey by Knowledge@Wharton and SAP's Performance Benchmarking team reveals that while the hype and excitement surrounding cloud computing is reaching a fever pitch, many businesses are still expressing concerns over cloud security and IT integration issues. The survey also shows that while many people agree that the cloud is revolutionizing business, they still do not fully understand how it works.
How will these tensions surrounding the cloud be resolved? How will the cloud transform businesses in the future? What kinds of benefits will the cloud bring, and is it worthy of the current hype? Knowledge@Wharton discussed those questions and the survey results with David Spencer, vice president at SAP, and Don Huesman, managing director at the Wharton Innovation Group.
An edited transcript of the conversation appears below.
Knowledge@Wharton: Knowledge@Wharton and SAP's Performance Benchmarking team just conducted a survey about the business value of cloud computing. The survey found that 85% of respondents believe that cloud computing will transform business. Could you speak to us about how cloud computing can make a company more competitive by leading to measurable efficiencies and improving business processes?
David Spencer: There are three areas specifically where the cloud can help our customers be more competitive. One: The Cloud gives individuals better access to areas of their business that the traditional business model previously didn't give them access to. Two: The speed at which you can deploy a solution gives cloud users an opportunity to realize revenue or benefits much more quickly. Three: The cloud is delivered in a way that makes the total cost of ownership much better. Also, in my opinion, the user interface is more flexible, and it is much easier to deploy.
Knowledge@Wharton: Could you offer any examples of companies where you see this happening?
Spencer: I'll give you an example of a customer that just recently deployed our Customer OnDemand solution and did it in two weeks. The customer came back to us and said, 'If we had done this using a traditional model, it would have taken us six months to a year. You've done this in two weeks!' The deployment was completed faster than any other traditional model that was out there. As for the cost -- because it's a variable cost versus a fixed cost, they were able to budget the cost over multiple years instead of taking all the costs upfront.
Knowledge@Wharton: Don, any comments?
Don Huesman: From the Wharton side, I can tell you that the cloud has provided an opportunity for entrepreneurial-minded students with new ventures to establish fairly sophisticated capabilities quite quickly and on a relatively low start-up budget. The cloud allows them to quickly enter a marketplace or put together a platform for delivering a new service at speeds that were not possible before the existence of cloud services. The cloud has certainly provided agility for these entrepreneurs. It creates a more level playing field, and thereby generates more competition.
For example, ElectNext is an enterprise that was established to solve the problem in this country of having an essentially broken political system. By using the cloud, the organizers could efficiently respond to the level of site usage and the level of infrastructure that they needed at any one time. They added on new servers when they found that service was slowing as a result of inadequate resources. In the past, this wasn't possible. You might have needed to make a guesstimate about possible incoming site traffic and acquired hardware and talented software people to handle this. Now you can respond almost in real time to changes in site demand.
Knowledge@Wharton: Back to you, Dave. Some 65% of survey respondents said that cloud computing would amplify other technology megatrends. Could you tell us which megatrends will be amplified, and how this might happen?
Spencer: Yes. Let's talk about mobile, social media and big data. We're seeing more data coming into companies. For example, people are constantly talking about company interactions and opinions on Twitter, Facebook and other sites. Now companies have a massive amount of data that need to be analyzed. One implication that I'm seeing is that companies have to take a more holistic view of their branding. They have unprecedented access to their customers via social media, and individual customers can heavily influence other customers. Now companies can individually market themselves based upon market trends that are going on.
I'll give you an example of where I'm going with this. Recently, the NFL decided to use SAP's cloud solutions to run their Fantasy Football service. That's accessing four million users. The NFL's IT division may not have had the infrastructure designed to support that before, but because of cloud computing, big data and social trends, they now have an opportunity to market individually to people.
Knowledge@Wharton: Don, do any examples of megatrends come to your mind?
Huesman: One megatrend that is central to cloud computing has been referred to in the press as the consumerization of IT. Before I go further, I would like to note that cloud services have been around a long time. They are a 20th century innovation. Gartner has just indicated that we're currently at the top of the "hype cycle" for the concept of cloud computing. I think it's related to the consumerization forces. So, Drop Box and other cloud services for e-mail have enabled both individuals and businesses to take control of their own IT infrastructure needs. I think this force of consumerization has been a real driver in why cloud computing is so interesting now.
Spencer: Adding onto that, the cloud was originally designed for small companies. It all started with non-strategic applications for smaller companies. But over the last couple of years, we have seen larger companies start to see the model and understand how it fits into their business. Now what you're seeing is even larger companies running their entire HR systems in the cloud. Larger companies are, without a doubt, adopting cloud solutions more than ever.
Knowledge@Wharton: Related to your comment, our survey found that 66% of respondents said that cloud computing is helping their companies create new, strategic options. HR is one of those options. Are there any other strategic options that cloud computing opens up for companies? What value does that create for the users?
Spencer: HR and sales automation are the two that have gotten a lot of attention. But I'll go back to what Don said, too -- it's also the commercialization piece of this puzzle. Companies are now tailoring marketing directly to customers, who influence other customers.
Knowledge@Wharton: Don, what do you think about the strategic options?
Huesman: From the point of view of students and start-ups, the consumerization of IT is an important component. The cloud provides application software that can be used on hand-held devices -- such as iPads and Android -- and this allows a mix of vendors to provide very specific types of application software through an app store. I see this as an opportunity to change the market for traditional enterprise services by broadening the vendor base and increasing competition. Also, I should point out that in an app store environment, and as a part of the consumerization of IT, there will be downward pressure on pricing.
Knowledge@Wharton: Both of you have spelled out of the value of cloud computing. But one of the counterintuitive findings in our survey showed that while many people believe that cloud computing is important, they don't seem to understand much about it. Don, why do you think this is the case?
Huesman: I just mentioned the Gartner "hype cycle." Gartner showed 2012 as the time when the hype slipped off the peak and headed towards "the trough of disillusionment," which they think is a necessary phase for any type of effective technology before it reaches what they refer to as "the plateau of productivity." And Gartner is predicting that to happen with cloud computing over the next two to five years.
Related to this "hype cycle," I think that there is a tendency, particularly in the United States, to look for technical solutions to multi-dimensional problems and get very excited about the solutions, until we begin to see some limitations. Then we get very disappointed, until we eventually discover that there is still a decent amount of utility to take advantage of in these technical solutions. So, in this case, I don't know that the first-mover advantage in this space is very strong because I think close followers can also take advantage of the new capabilities that are emerging.
Knowledge@Wharton: Dave, what's your view?
Spencer: I think one of the reasons there is general confusion has to do with the fact that there are different definitions that are used when referring to cloud computing. For example, there is SaaS (software as service), there is a virtualization of an environment, there are also hosting environments. This leads to confusion because there are different kinds of cloud solutions. In fact, at SAP we try to describe it in terms of value across our customers' entire enterprise, and even beyond. We talk about our cloud portfolio consisting of four main elements: apps, cloud platform, infrastructure/lifecycle management, for investment protection, and finally, the business network, which refers to networks of buyers and sellers doing business in entirely new, social ways, all enabled by cloud computing.
Knowledge@Wharton: Interesting. Perhaps this lack of understanding could be related to this finding from our survey: 18% of companies said they had extensively or completely adopted cloud computing, but half the respondents had minimal or no adoption. Could this general misunderstanding about cloud computing be related to the fact that not many people are using it?
Spencer: If I look at the customers that I represent, most have adopted some kind of cloud solution or at least have looked at it. Now, they may not have deployed it 100% because they may have started out using the cloud solution in smaller organizations or smaller businesses processes, but every day I see people looking at deploying some kind of cloud solution.
I think the other factor is that if you're looking at larger organizations, they would have already made sizeable investments in their IT infrastructure, so quickly moving to a cloud solution may not be the best investment for them right now. What a lot of companies are looking at is called the hybrid solution, where they are taking select business processes and putting them into the cloud, while still leveraging their on-premise IT infrastructure at the same time.
Huesman: Dave has touched on an important point here. I think there is often resistance on the part of IT departments in some companies towards the concept of the cloud. This resistance is based on concerns about security, reliability, dependability and the robustness of the solutions that are being offered. But I think all those concerns are overstated. The current resistance reminds me of when the personal computer was introduced. Large companies resisted its deployment for a very long time until personal computers were being purchased by CEOs. At that point, the IT organizations had to figure out a way to integrate them successfully. I think we're seeing a similar dynamic now where there is resistance for legitimate concerns, but that's being trumped by consumerization.
Knowledge@Wharton: Do you expect to see CEOs in the cloud soon?
Huesman: Oh, I've seen them there every day.
Spencer: On my side, I can say that we run our business in the cloud 100%. Every single business process in the SAP cloud business unit, from travel to HR, is being run in the cloud.
Knowledge@Wharton: Dave, for those who are resisting cloud solutions, some of the biggest concerns seem to be around security and integration. In fact, 67% of survey respondents identified those two issues as the ones that concerned them the most. How do you deal with this?
Spencer: We try to address the security issue head-on to understand our customers' exact concerns. Now, there are different privacy laws in different countries, which we've been able to satisfy. But there are some companies and government agencies that have very strict policies around what can be run on the cloud and what cannot. So, we try to address these issues. But at the end of the day, it really comes down to that fact that some people are just going to be cloud laggards. For those people who are going to be at the back-end of the curve in deployments, you just have to make sure they are comfortable when they're ready.
Knowledge@Wharton: What do you think, Don?
Huesman: We were one of the first organizations at Penn to attempt to outsource our student email to a cloud service provided by Microsoft, but we had a bad experience early on. It was so bad that we had to move back to our previous system. That early mover experience leaves people nervous about making the move again. But now we're at a point where outsourcing e-mail is something that can be done in a more reliable fashion.
These issues make IT directors and IT workers believe that it's better to work through the late night hours to repair and reconfigure hardware and software services rather than wait on the phone anxiously hoping that someone else in another distant city is fixing the issue. It's a different position to be in psychologically. It may also be a challenge for IT employees to be able to work in partnership with a cloud services company. Not all people will be ready for this shift. It's certainly a period of transition in the IT industry.
Knowledge@Wharton: Dave, do you encounter some of the issues that Don just described? How do you deal with them?
Spencer: We deal with these issues every single day. We look at this as a true partnership: The clients working with SAP have to feel comfortable that they're working with a responsive business that is going to continue to deliver innovation and improve security. Our customers want a business partner that can handle all aspects of the cloud. For example, SAP has to proactively deal with issues related to deploying business applications on personal mobile devices and controlling the data that's out there. That's one area where SAP is very advanced.
Knowledge@Wharton: When it comes to hybrid solutions, are there some areas where it's easier to move things into the cloud than others? This relates to one more finding in our survey where 75% of the participants said they thought cloud computing meets minimal business needs. Could you speak to that?
Spencer: Well, if you take SAP as an example, we offer cloud applications in four big areas, focusing on the parts of our customers' business that are most important to them: people, money, suppliers and customers. We're also integrating social aspects into these cloud services. For example, we offer collaboration rooms where teams can work together to address customers' needs. We build the analytics and we also build in the mobile applications in these systems.
The reason I mention this is because not all customers want to use the entire suite of cloud offerings. They may just want to take one application, such as expense reporting, and put it in the cloud. But then this service has to be integrated with other business applications that are not in the cloud.
Knowledge@Wharton: Don, do have anything to add?
Huesman: There are some types of document storage services that are easily moved to the cloud that might fall under the "minimal business needs" category. While some businesses in health care and law, for example, may not be able to use this kind of service due to strict regulations, other companies in other industries may find that this kind of document management system is easily moved to the cloud. This service can free up a lot of internal IT resources and can be quite beneficial in terms of savings and redeploying talent to tackle other issues within the organization.
Knowledge@Wharton: I'd like to end with some questions related to the future of cloud computing. Nearly 70% of survey respondents said that they believe cloud computing will lead to major changes in their company's performance over the next five years. What changes do you envision and how will they come about?
Spencer: With the cloud, we are seeing people deploy technology faster and realize and recognize revenue and benefits faster. The cloud enables companies to access their customers in a way that is faster, more dynamic, more flexible and less expensive, and the system is integrated with existing infrastructure. People are also willing to make investments in cloud computing because it offers a flexible payment cycle. Furthermore, what may have taken years to build beforehand can now be done in months.
Huesman: The possible future that I see emerging is an extension of what has happened in the music industry with the development of iTunes. This classic cloud service disrupted and changed the dynamic for the music infrastructure providers. It extended music to a much larger number of providers, and it disintermediated some of the traditional businesses that controlled global music distribution. I wonder what that might look like in the space of productivity suites and enterprise resource packages. Is it possible we might see the emergence of an ecosystem of products that would allow for much broader competition? As I mentioned earlier, I think the cloud will contribute to downward pricing pressure.
Knowledge@Wharton: Which industries are ripe for that kind of disruption, thanks to cloud computing?
Huesman: I think education is in a position to finally be affected by the Internet. The classroom of today looks very much like it did in the 19th century, but I think that's beginning to change. The change is partially due to cloud-based services, which allow the highest-quality classroom curricula to be compiled together and offered to organizations for free.
In terms of business, I think IT will become more streamlined and become more of a commodity service, which will cause downward pressure on IT expenses.
Knowledge@Wharton: That's very interesting. Nearly 60% of our respondents believe that cloud computing will bring about precisely that kind of cost saving and encourage more business agility. Now I'd like to end by discussing how we might climb down from the "mountain of hype," as Don said earlier, to the "plateau of productivity?" How can this be implemented in a way that delivers real value to companies?
Huesman: I've often been involved in projects where people hoped that an IT investment or a change in IT infrastructure would generate significant savings. But I have seldom seen that happen. However, I do see instances when deploying a new technology has provided businesses with an agility that makes them more competitive. That changes the game. While the focus on cost savings is very attractive to senior management, I would instead encourage them to focus on how the cloud can increase their agility to produce new services and new products of value for their customers.
Spencer: Let's look at three things: agility, speed and investment in technology. In terms of agility, we see the cloud as being able to quickly deploy processes that are integrated with current systems. In terms of speed, you can deploy cloud processes so much quicker that traditional processes. From an investment standpoint, people are more willing to take on smaller cloud deployments and then ramp up and build on that success as they go.
Knowledge@Wharton: Do you agree that the opportunity to increase agility is higher than the ability to bring about cost savings?
Huesman: That's been my experience in working with companies.
Spencer: I think it's a combination of both.
The Penny Pinchers of the Cloud
Cloud computing has made Web startups cheaper than ever to run. Some organizations are now learning anew how to be budget-conscious.
Cloud computing (see “Who Coined Cloud Computing?”) provides a way to scale such big, complex projects quickly, but figuring out how much to spend can be a balancing act. Reed, for example, couldn’t afford to let any websites, e-mail programs, or Facebook apps crash, nor did he did want product engineers worrying about computing costs. But neither did he have cash to burn renting excess server capacity.
He turned to Cloudability, a startup based in Portland, Oregon. The company’s analytics software, he says, was suggested by Amazon Web Services, where the campaign was renting more and more server capacity by the month. At that time, Amazon’s cloud computing business did not have particularly good tools for helping customers track their spending, let alone figure out how to rein it in.
Cloudability is one of a crop of startups aimed at helping businesses understand the costs of cloud computing as the technology matures and much larger businesses get more comfortable with the idea of renting or leasing remote computing capacity to run their software and IT infrastructure.
At many companies, the “cloud” can involve an unwieldy, decentralized mess of hundreds of individual accounts and credit cards that can be hard to track. “Getting a developer to turn off a cloud server is like getting a kid to turn off the light switch,” says J.R. Storment, Cloudability’s cofounder and chief customer officer. Often they may just forget.
His startup, founded in early 2011, says it has signed up 4,500 customers that account for a total of $260 million in cloud spending, or, as Cloudability estimates, 10 percent of Amazon Web Services’ usage. Startups like Newvem and Cedexis, and Amazon itself, now aim to provide a way to monitor cloud usage more effectively as well.
The company’s basic software is free, but there’s also a paid version with more features. It tracks a business’s cloud billing according to up to 30 different metrics, such as time period or region. It can also be used to break down spending by project or by application—a key tool for developers who are working to, say, write more efficient code.
Storment believes such tools will help Fortune 100 businesses feel more comfortable with the cloud’s on-demand rental model, which could save them money in the longer term. The software can also protect against accidental account overages when developers forget to switch off cloud accounts or services. It can also warn of hacking—Storment says one customer found out when it got alerts that its bills were abnormally high.
Reed and a few other top-level technology managers at the Obama campaign had Cloudability’s reports e-mailed to them, and they were able to use them to optimize the amount they were spending on Amazon’s Dynamo database, a high-performance service that is good for storing Facebook messages to send back and forth quickly but generally too expensive for other uses.
The Obama campaign is an extreme case, but for many startups, there comes a time when the days of freewheeling, scrambling growth need a reality check. Maybe someone was hired to pay attention to costs, as happened when Pinterest, a social network that was one of the fastest-growing sites of all time for a few months, hired Ryan Park to lead its technical operations in 2011.
“Our number one priority was keeping up with the growth,” he says. But “one of the first things I did was to figure out what we were spending money on, and on what vendors and why.”