Thomas Bittman

Gartner analysts answer thousands of questions daily on various IT topics, and we learn a lot about IT trends by talking to our clients. For example, for the last twelve months we answered more than 1,000 inquiries from clients specifically on the topic of cloud computing. Which end-user verticals are asking the most questions?

1. Financial services (12%)
2. Manufacturing (10%)
3. Business and management services (10%)
4. Telecommunications and equipment (9%)
5. Government (7%)
6. Insurance (6%)
7. Oil, gas and electric (5%)
8. Professional/specialized services (5%)
9. Schools and education services (4%)
10. Food (4%)
11. Retail (4%)
12. Healthcare (4%)
13. Media (3%)
14. Chemical and pharmaceutical (3%)
15. Military and National Security (3%)
16. Freight services (2%)
17. Energy management (2%)
18. Membership organizations (2%)
19. Commercial physical research (1%)
20. Other (4%)

At this week’s Gartner Symposium in Orlando, there was a noticeable shift in the end user discussions regarding virtualization and cloud computing, and a few surprises:

1) In my presentation on server virtualization on Monday, before I started, I asked the audience how many of them considered private cloud computing to be a core strategy of theirs. 75% raised their hands (I expected maybe one-third). Clearly, everyone has a different idea of what private cloud computing means (or doesn’t), but the fact that so many people have glommed onto the term is very interesting.

I described the three most common things that were being described as private clouds:

• IT defending its turf: Shared services that were being re-labelled as private clouds (but without a self-service interface, or much automation at all)
• Vendors defending their products: Old products being re-labelled as private clouds in a box (I described most of these as “lipstick on a pig”)
• Advanced server virtualization deployments: Although few have a true self-service interface, the intention is certainly there

2) My one-on-ones shifted heavily from virtualization toward cloud computing and private cloud computing. I had 18 one-on-ones that discussed server virtualization, and 26 that discussed cloud and private cloud.

3) Several one-on-ones were very interesting: one company with more than 5,000 VMs heading toward 10,000; maybe a dozen that were focused entirely on taking their leading-edge virtualization infrastructure into private cloud; three companies considering discussions with Terremark on possibly licensing their service interface and usage metering software (they know VMware will probably deliver something in a year, but don’t want to wait); several users wondering what Microsoft’s virtualization strategy was, and especially whether they were focused on private cloud computing as a next step; one that questioned VMware’s vendor maturity with respect to customer relationships (felt that Microsoft did an excellent job with TAMs and so forth to truly work with the customer as a trusted advisor – while VMware’s TAMs were more like communications conduits).

4) In my “debate” with analyst Eric Knipp (on the topic of public vs. private cloud computing), we took a vote at the beginning and end of the session to test audience opinions on the subject. At the beginning, about 15% felt that public cloud computing would “win” (the question was intentionally vague to let the attendee decide what “win” meant). About two-thirds though that private cloud computing would win. By the end of the debate, perhaps a few more thought public cloud would win, about 40% thought private cloud, and about 40% voted otherwise (which included a hybrid model). I suppose the surprise to me (yet again – do I never learn?) was the overall momentum in the concept of private cloud.

The challenge with private cloud computing, of course, is to dispel the vendor hype and the IT protectionism that is hiding there, and to ensure the concept is being used in the right way – as a stepping-stone to public cloud, based on a timing window, the lack of a mature public cloud alternative and a good business case to invest internally.

Tomorrow I leave for Orlando for our annual all-up IT conference of the year in Orlando, Florida. Attendees see this conference as a one-stop week-long update on all IT industry trends, an opportunity to network with thousands of their peers (including more than a thousand CIOs), and an opportunity to have face-to-face one-on-ones (as opposed to phone inquiries) with analysts on tough problems that need actionable advice. And maybe have one last week in some warm weather. Symposium delivers all that. But it delivers something different for me.

Gartner analysts spend tremendous time and effort preparing for this conference. Presentations are due to our editing and multimedia departments weeks in advance, but for some reason, IT doesn’t stop, so there are very few analysts who turn in their work on time. Personally, I always make changes as late as the day of the presentation – that’s just a little more real-world.

I love these conferences. During the year, I spend a large percentage of my time on the phone with clients (600 or so calls this year?). I also visit with clients face-to-face throughout the year (I think I visited with perhaps a hundred this year). However, nothing compares with the density of client conversations that take place at Symposium.

For me, Symposium is about four days of constant client interaction. This year, I’ll deliver two presentations (one on cloud and private cloud computing, one on virtualization), a debate (is private cloud real?), a client roundtable, about 40 one-on-ones, two breakfasts with clients, two lunches with clients, a dinner with one client, and another dinner with a few dozen key CIOs. History says, all remaining open time will disappear as soon as I arrive. This will be solid 7am to 10pm client discussions.

I know they come to get advice from us, but we come to hear about their issues.

There is no better way to measure the pulse of IT users. Every interaction is a specific issue that a decision-maker needs to solve. I’m put on the spot, every thirty minutes or so. Sometimes, I can’t answer a question, but I can usually move them closer to an answer, or line up another Gartner analyst who can. But most of the time, there’s a two-way transfer that takes place – I give the client actionable advice based on 25 years of experience and thousands of real client interactions, and they give me a new and up-to-date real-world issue.

An analyst’s job is about being an expert who can help clients with actionable and strategic advice. But an analyst’s job is also about being a learning machine. There’s a tremendous amount of information out there, and it’s our job to find the important nuggets, filter out the chaff, find the patterns, and determine the actions that help our clients the most. And this is what makes the job interesting and keeps me on top of my game.

This year I’ll be active tweeting during the conference. Of course, nothing confidential about individual clients, but I’ll tweet about the pulse of the market and things that are coming up often (tombitt on Twitter, and I’ll hash my tweets with #GartnerSym).

And for those of you coming to the conference, I’ll see you there, and hope you get as much out of it as I do!

It seems like some hard lines have been drawn in the market over cloud computing versus on-premises computing. On the one hand, the proponents of cloud computing are promoting a massive shift of software development toward cloud platforms, designing for multitenancy and massive scale. No more software packages – just buy software as a service. How can you possibly compete on price with service providers leveraging huge economies of scale?

On the other hand, those defending traditional IT are pointing out the many and glaring flaws in today’s cloud computing services, and lining up behind the private cloud computing bulwark. IT vendors interested in maintaining the enterprise business they enjoy today are also co-opting the private cloud computing concept as only a slight modification of what enterprises have been doing all along.

This cloud computing argument is a lot like belly buttons. You’re either an innie or an outie. The unfortunate thing is this has led to some serious investment in navel-gazing.

Perhaps a dose of pragmatism would help. Or lint removal?

The cloud computing services needed to deliver the majority of IT services needed by customers do not yet exist. There are limited SaaS offerings today, service-level requirements can’t always be met, glaring security holes exist, regulatory compliance requirements haven’t caught up with technological capability, cloud providers tend to be proprietary and monolithic – just another opportunity for lock-in.

On the other hand, private cloud computing services cannot have the economies of scale that many large providers will enjoy. The complexity and speed of technology change will be hard for any internal IT organization to handle, especially smaller ones. The investment needed to build a private cloud service may be immense, and the resulting architecture could be a dead-end.

Gartner believes that there are quite a few services available today from cloud computing providers that are ready, and cloud computing will gradually fill more and more computing service needs. Where opportunities exist for a business model, cloud providers will fill those gaps – even when the number of potential customers for a service range in the hundreds, rather than the millions. Brokers and interoperability standards will emerge. SLA and security guarantees (for a price) will evolve. And don’t forget the completely new services that will emerge because of cloud computing and scale. It’s coming.

But it may take many years for some services. So while it is evolving, private cloud computing may make sense. It’s a question of ROI, not religion (or belly button architecture). So how you build your private cloud services matters – it’s important to build it with relatively rapid return on investment, in a way that eases a migration to external cloud computing at some point in the future. That point could be very far away. Or a hybrid model (both private and public) for some services might make the most sense in the future.

And there will be some services where cloud computing won’t ever make sense – perhaps there just won’t be a business model for a service provider, or perhaps it is simply too customized for a specific business – a real differentiator. Or perhaps it changes often, and doesn’t easily fit a concept of standard and relatively static interfaces. I think this kind of service will be a minority, but it will exist.

So hopefully we can raise our eyes from our navels, and make intelligent business decisions that will embrace the cloud computing concept as a part of our toolbox, and build a rational evolution to take advantage of cloud where and when it makes sense. And cents.

I need to rant a little.

As an analyst at Gartner, I can’t describe how angry I get when I read bloggers spouting as “fact” their opinion that I and my teammates have no integrity. That we can be “bought.”

In my 14+ years at Gartner, I have never, ever allowed a vendor to influence my opinion with anything but facts. Period. They have certainly tried to influence me with non-facts. I can say this definitively – it has never worked.

I don’t think there is a single vendor that I have dealt with who has not been very angry with me at some point. Tough. I’ve been yelled at by many IT executives – including the CEOs of Microsoft and HP, and many other firms. I can’t think of one of those cases when I changed my analysis one bit. I can’t speak for other firms, but at Gartner, getting yelled at by a CEO is a badge of honor. Being proven right as time goes on – priceless. 

I certainly spend time helping vendors with their strategies and their marketing messages – and I enjoy doing it. Frankly, the ones who yell at us the most seem to respect our opinion the most. We can spot holes a mile away, and engaged early enough, we can help vendors fill those holes with real product offerings – that not only help the vendors, but help our end user clients. And my primary business is helping end users.

I worked at IBM for 11 years. During the last few years, I was doing some technical evangelizing to analyst firms – and hiring these firms to write white papers. There were plenty available for hire – and still are. There was only one firm that wouldn’t allow us to edit their work – Gartner. Only one. We tried, but they wouldn’t budge. They could not be bought. It was unusual, and I completely respected that. When the opportunity came to join Gartner, I jumped. I wouldn’t have even considered anyone else.

Since I’ve joined the firm, my appreciation for the integrity of the firm and its analysts has only increased. If I felt that the firm’s integrity was bending, first I would fight it with all my might, and then I would be outta there. I’m not gone yet!

So I understand the impression in the marketplace that analyst firms can be bought. But that’s not where I work. My integrity is very important to me. I’m sure we’ll continue to make enemies of vendors, and bloggers who have a vested interest in one thing or another. Badge of honor! But my goal is to provide value to my clients, and to be proven right over time – priceless!

End of rant.

I was discussing cloud computing with a client today, trying to find a good market analogy to position how cloud computing would evolve.

The electricity utility example doesn’t work. Neither does the water utility. There are two reasons: (1) Computing is a rapidly evolving technology, and (2) Service requirements vary widely for computing. Electricity production and distribution hasn’t evolved much since the invention of AC that made distance distribution possible. How many forms of water are needed around the world? It’s H2O – maybe it can be potable, purified, or come at a special temperature, but it’s still pretty basic stuff.

Analogies are never perfect, but done well, they can be illustrative. I think the best analogy for cloud computing services is transmitted music. Bear with me.

Music was first transmitted over radio waves 93 years ago. At the time, phonograph cylinders were used for music storage. There were probably pundits at the time who could look ahead and declare that individuals wouldn’t own their own copies of music – the radio would provide music for them “as a service.” Owning piles of phonograph cylinders, managing them – too expensive, too complex.

And in 93 years, we have gone from AM to FM to subscription-based satellite radio. It’s gotten better and better. The choices have grown substantially. The quality has improved dramatically. But so has on-premises music. The phonograph record, eight-track tapes (OK, not so great – but easier than lifting a needle to find a song), CDs, digitally-encoded music. Why hasn’t on-premises music died?

(1) Technology has continued to evolve both for off-premises and on-premises delivery of music. There is an arms race – while choices and quality grows off-premises, it also grows on-premises.

(2) Quality of service (and choice) matters. Yes, I can turn the tuning dial, and choose a different genre on Sirius, but if I want a specific song, now, and on a plane or in a cave or while I’m exercising, an Ipod delivers it. I’m willing to pay more to ensure I have choice and quality.

The bar for when off-premises vs. on-premises is the right answer will constantly change – there are many use cases where transmitted music services are good enough – and maybe better than I can afford myself. But there continue to be use cases where only on-premises can deliver.

And the same will be true with cloud computing services. Yes, a provider can aggregate users and drive tremendous economies of scale. Yes, managing your own IT is complex. But the same technologies that are driving and evolving to make cloud computing more automated and efficient are available (at a different scale) to end users. More importantly, service requirements can vary widely. Some services will naturally move to the cloud, and new services will evolve in the cloud, quickly. Other services will remain on-premises for quite some time (for differentiating, or service, or regulatory, or customization, or security, etc. reasons).

Will they eventually move to the cloud? I think that is very hard to answer, and maybe not important. The music analogy shows an example where the inevitable wasn’t quite so inevitable. The challenge is to find the right balance point, and try to predict how that balance point will move over time to ensure you get a good return on your on-premises investment. Predicting the next five years is important. And for the next five years, there will be many reasons to invest in private cloud services – with an eye on what’s coming in the cloud.

There’s a big change coming to the server virtualization market, and it has serious ramifications.

The midmarket is waking up. At Gartner, the number of midmarket clients who are just starting out with virtualization has been growing exponentially for the last year or so. We just completed a worldwide survey of nearly 1,500 companies with 100-999 employees. What you see is pretty compelling evidence that a market that nearly ignored server virtualization for years started to wake up about two years ago, and is rapidly becoming a dominant opportunity for vendors.

Furthermore, it’s clear that midmarket enterprises virtualize at a different rate than large enterprises. Large enterprises virtualize over a long period of time, at the rate of hardware replacement. Midmarket enterprises tend to virtualize as a part of a single project, or over a 1-2 year period. 

We predict that midmarket enterprises will have a higher percentage of servers virtualized by year-end 2010 than the global 500 – who have been virtualizing since 2001. That means 2009 and 2010 will be huge.

What does this mean? I think it means that VMware messed up, and effectively ignored the midmarket (especially in terms of entry price) until Microsoft came to the table. I think the price war is going to take a serious toll on vendors at the low-end. I think midmarket software vendors who have been able to ignore virtualization are going to have to make changes very quickly – in terms of support, but also pricing and licensing. I also think that we are going to have some pretty messed up midmarket enterprises – who virtualize too fast, use immature technology, lose some control of their environment, open new security gaps, etc.

Gonna be a bumpy ride! 

Few organizations that I talk to seem to understand the strategic ramifications of server virtualization. They tend to think about cost-cutting – virtualization simply as a form of efficient consolidation. We’ve surveyed our clients – those starting out on virtualization say they are doing it to save money. They are thinking tactically.

Hey, there’s nothing wrong with saving money, but strategically, virtualization is not primarily about cost-cutting. Strategically, virtualization leads inexorably down a path toward flexible sourcing, and cloud computing.

Even our surveys show that organizations who are well on their way toward virtualization change their points of view – flexibility, agility, speed move to the top of the list.

What is virtualization doing to these people? There are at least five things that virtualization does to unlock the door to cloud computing, and push organizations faster in that direction:

1) Enables economies of scale: This is one way cloud providers squeeze their costs in order to make money. Enterprises can do it too!

2) Decouples users from implementation: It’s amazing to me how many business units are closet server huggers. They like to stipulate how their software is deployed. They like to know where the server is located. They don’t like to share! Virtualization forces the relationship to change from a specific implementation, to service level agreements. It also makes it possible to choose alternate sourcing – because if the customer relationship is services, IT can choose how the implementation is sourced.

3) Speed, flexibility, agility: Early adopters of cloud computing talk about how quickly they can get new servers online. Compared to the 4-6 weeks it takes an average IT shop to deploy a server, just about anything is faster. However, virtual machines can be deployed roughly 30 times faster. It doesn’t take a cloud to improve speed. And, of course, operational processes and management tools need to change to deal with speed. And speed changes business expectations and behavior – it changes culture.

4) Breaks software pricing and licensing: You can’t charge users for physical capacity when only a small portion of that is used. You can’t charge users for every potential server the software might be running on. You’ve got to charge and license based on some kind of usage model. Of course, you can charge whatever you want until users get smart, but change is inevitable.

5) Enables, motivates chargeback: When servers can be delivered in minutes rather than weeks, IT users ask for more – roughly two times as much, based on feedback from our clients. The natural barrier is gone. Unless there is a cost, a friction, associated with a server deployment, how do we make good business decisions? IT needs to focus more on usage accounting, and chargeback is growing as a mechanism to manage virtual capacity usage.

Economies of scale, shifting users to a services-oriented relationship, delivering much faster, forcing software prices to align with usage, charging business units based on usage. Sounds like cloud computing to me.

Virtualization, private cloud, cloud – that’s the natural evolution.

The thing is, most major IT vendors (not all) get this. Why do they care? The time to influence an enterprise’s cloud choices, software architectures, management architectures for the cloud, standards, are when organizations are virtualizing. So is cloud computing on your strategic plan? It may not be on yours, but it certainly is the plan the major virtualization vendors have for you! Be proactive, take charge of your own on-ramp to the cloud, or get taken!

Summer’s almost over – time for less sun and more clouds!

I’ve gotten a lot of client questions recently about how to get started building their private clouds. Their vendors are at their doorstep hawking private cloud computing. Clients are asking about what technologies to use, which vendors to choose, how to build one, etc. I say, “Whoa!”

This is no field of dreams. “If you build it, he will come” is not a reasonable strategy for private clouds.

We’ve got to get our IT people to stop thinking about products and technologies and even architectures first, and instead to focus on understanding their service requirements first.

Start by understanding your service catalog (most organizations don’t have one), understand the SLAs and costs for each service (most don’t know that, either), build strategic plans for each of those services (does anyone have this?), determine which ones might go to the cloud in the future and when that cloud service will be “ready” (OK, this takes some work), make your ROI decision about building a private cloud service, and then you can start looking at architectures and products and technologies.

There has been an awful lot of definitional talk about cloud computing and private clouds for the past year. Time to move on. My cloud research at Gartner is focusing on case studies, roadmaps, and best practices for private cloud computing – and debunking the incredible amount of vendor hype that is hiding some real value in this trend.

I’m sure I could come up with more than four, but let’s throw a few out there that I continue to hear from people trying to understand the phenomenon. Don’t misunderstand my intention here – I believe cloud computing will be huge, especially for commodity services, especially for small businesses and start-ups, especially for new and innovative applications that will leverage massive scale and low barrier to entry. But…

Myth: There will be a “big switch”
Fact: There will be a slow migration (including development of private cloud services), the migration will take decades, and even then quite a bit of IT will stay in-house; in fact, most of the interesting stuff will be hybrid models, long-term.

The electricity analogy really doesn’t fit well. Unlike electricity distribution (using AC instead of on-premises DC), IT is evolving at an extremely rapid rate. The number of enterprises generating their own electricity in 1887 was miniscule compared to the number of enterprises generating their own IT today. Even so, the electricity grid did not take place over virtually overnight – it took more than two decades before centralized utilities produced more than half of the electrical production in the U.S.

In many ways, the invention of AC was like the invention of the Internet. So why hasn’t computing across the Internet replaced enterprise computing yet? In IT, the “distribution” mechanism has been in existence for more than a decade. What’s really changed are technologies that enable economies of scale and sharing. We should not ignore the fact that enterprises can gain some economies of scale themselves, internally, for example, through virtualization – and enterprises aren’t ignoring it. The evolution toward cloud computing is a multi-variable equation, and does not have an inevitable conversion of everything to cloud services.

Myth: Cloud computing is just an evolution of “fill in the blank”
Fact: Cloud computing did not appear out of nowhere. Some say it is just the next version of outsourcing. Some say the next version of the web. Some say the natural evolution of virtualization. I say it’s all the above. The web created the standards and connectivity needed to make cloud computing possible. But, economies of scale do not occur unless you have technologies at the back-end that enable efficient technology sharing – multitenant applications, virtual machines, parallel programming mechanisms, automation, etc. Sprinkle in a growing demand for speed in the marketplace, and the industrialization of IT (including increased commoditization of hardware and open source), and cloud computing – has been evolving for years. And still has a ways to go.

Myth: Only megaproviders will win
Fact: There are diminishing returns to economies of scale, there are many fragmented markets that have good enough scale for smaller providers, and innovation makes provider agility a critical offsetter to size. We’re not going to have a handful of megaproviders, we’re going to have thousands of providers, and it will be very Darwinian.

Myth: Cloud computing is about IT commoditization
Fact: While services offered in the cloud may be commoditizing, the usage of those services may not – new, innovative businesses, proprietary analysis of data in the cloud, etc. – new applications matter. In fact, innovative use of cloud computing services will be be a huge reason why IT does matter, and innovative use of IT will remain a critical business differentiator.

And be careful – cloud computing services are not always cheaper. Providers gotta make a living. Amazon’s recent introduction of Reserved Instances was both to help Amazon plan and manage capacity, and to lower the price to compare better when workloads are not dramatically elastic. There’s a reason that some startups born on Amazon created their own data centers as they got bigger and their businesses became more predictable.