Jon Hansen’s Cloud Computing in the SaaS World
Posted by Bob Warfield on May 1, 2009
Jon Hansen runs an excellent podcasting/Internet radio show called PI Window on Business. Recently I was invited to join one of these casts, but couldn’t due to another commitment, but I wanted to pass along this program because it concerns Cloud Computing and SaaS. Host Jon Hansen is talking with guest and blogger Michael Dunham from Scio Consulting. They’re talking about the recent McKinsey study that claims the Cloud does not deliver any savings for large organizations, and the program starts with a pretty decent introduction to Cloud Computing.
They key differentiator in Cloud Computing is that like SaaS, it is a Service. The customer pays for it without worrying how it works. The entire infrastructure in the background is transparent. That’s as it should be, BTW. There isn’t a lot of value and there is tremendous cost in having to be aware of every implementation detail to use a service. Dunham likens the Cloud to the telecommunications infrastructure that’s existed for a long time.
I wanted to go back over a number of issues raised in the show and give my own perspectives. This will be a longish post, because it was a half hour show that touched on a lot of issues.
There were lots of interesting parallels raised in the show. One theme is that Cloud really isn’t something completely new. As mentioned, we’ve had telecommunications and other kinds of Clouds for a long time. One of Hansen’s first questions was, “Who needs to be involved with the Cloud?” Will it be confined to an Oligarchy of Concentrated Expertise with the Large Players?
First thing to note is that the Cloud benefits from scale. It is essentially a commoditization phenomenon. There’s not a lot of benefit in buying services from a Cloud that is too small, unless those services are very unique. That will therefore drive scale on the vendor side of the equation except for more specialized kinds of Cloud. A lot of people I talk to wonder if Amazon isn’t already so far ahead of the pack scale-wise that it will be hard to catch them. The good news is that their offering is pretty generic, so there is an opportunity to differentiate. The bad news is that except for the largest possible companies, the IBM’s, Sun/Oracles, and the like, Amazon may already be too far along and so it will be essential for the other players to differentiate.
“Is it safe to say the expansion is occurring as the market is decentralizing? “
I don’t think of it so much as decentralizing as changing the locus of centralization. We move from centralizing around large corporate IT datacenters to centralizing around relatively fewer Cloud vendors’ large datacenters. One of the reasons companies are starting to scramble on the vendor side is this centralization.
The Cloud aggregates transactions. If I am selling servers pre-Cloud, I have lots and lots of customers. Win or lose any particular one, and it is not a big concern. There are a lot of fish in the sea. But, whichever vendor is lucky enough to close Amazon on their servers, wins a whole ton of virtual accounts (Amazon’s customers) by default. The stakes are much much higher. We will see Cloud providers dictating to such vendors the same way Wallmart and GE dictate to their suppliers how business will be done. Major new forces are being created in the market because Cloud Vendors represent the collective buying power of all their customers.
“So the general user and population don’t need to worry about this?”
The general user already spends most of their computing time in the Cloud. Every web app is in the Cloud from the standpoint of an end user. Don’t we already spend the majority of the time on our PC’s in web apps? So we have the reverse of the “last mile” problem of residential Internet access. The “last mile” is in place as we use all these web apps. The Cloud is about the “first mile” where the datacenter begins. What is your web browser connected to?
As the Enterprise grows increasingly distributed, this again favors the Cloud which is purpose-built for a webby world. It’s no accident one of the biggest Cloud vendors is also one of the oldest and most successful Web businesses. They know how to do that stuff!
“What about reliability, dependency on expertise, and support for the Cloud?”
Well, how reliable are your web apps? Do they crash more often than your Microsoft apps (LOL)? Mine sure don’t.
Are we happy with where the expertise lies with these apps? Tim Chou started the SaaS business at Oracle many years ago, and he is the first one I heard talking about the idea of “Who better?” Would you rather have your apps supported by your internal IT? Not me. They’re smart, but this is the first and only time they’ve ever run whatever app we’re talking about, and they didn’t build that app. I want the vendor to run the app for me and support it. They developed it, and they’ve run it for a lot of customers over a much longer period than my IT people. They’re the world’s foremost experts. Who better?
“What is the importance of standards?”
Michael Dunham was very concerned that we haven’t evolved enough standards yet in the Cloud world. I’m a lot less concerned. Amazon is a very straightforward service to adopt. The differences between what you have to know there are no greater than the differences between Sun SPARC Solaris vs HP/UX vs IBM AIX vs Wintel vs Linux Intel in a traditional data center. They’re no different than Oracle vs DB2 vs SQL Server vs whatever other platforms. In fact, they’re actually much less because a big part of what Amazon provides comes from these very same standards already. They haven’t added that much. I think it’s pretty straightforward to understand and take advantage of it. Standards are not holding us back to any appreciable degree, though IT loves to clamor for standards. It’s just their way of delegating some of the responsibility to understand.
What’s much harder for IT is the loss of control. They’ve built a lot of distinctive competencies over thousands of procurement decisions made over many years. They’re loath to revisit that fabric. But the advantages compel them to at least consider it.
“How is the organization involved with Cloud decisions? IT? Purchasing? Does this hasten the obsolescence of the CIO?”
There is a profound impact, no doubt about it. But I don’t think that impact is really that different from mega trends that have been at work in IT for a long time. IT has largely gotten away from being the arms and legs. They manage the arms and legs. I was having coffee with a friend from a Big SI the other day. She was lamenting they hadn’t yet embraced the Cloud, but she went on to say it was for the best because Customers lose control, there are security risks, and all the other chestnuts.
I responded that customers had already lost control and had all the same risks. Most of them are not running their IT today. The bulk of the people costs are going to outsourcers of one kind or another, either overseas, or IBM, or some other large service organization. I’ve worked with Fortune 500 companies that told me only 5-15% of the IT employees were actually employees of the company versus outsourcers. Why is the Cloud so different? She blinked, laughed, and agreed.
“Is there a concern that the big players don’t have enough experience with the Cloud? Remember, nobody ever got fired for buying IBM. Is this a big leveler?”
First, the big players are not blind to all of this. The Cloud and SaaS are highly disruptive. It’s very hard for them to flick a switch and be there overnight. The cost to their business model is just too high, and as was mentioned on the program, it touches every part of the organization and every aspect of doing business.
With that said, big license sales are slowing and have been for some time. Maintenance is becoming an increasing component. Acquisition of other company’s maintenance bases has become the growth vehicle. Ultimately, the source of organic growth will be Cloud/SaaS.
As I say, the big players are not blind. I wrote the post on the Red Cloud. I believe Oracle made the Sun acquisition largely in response to the whole Cloud movement. Moreover, Oracle has been active for many years with a SaaS business. It’s doing very well, though they don’t advertise it very loudly. SAP is less far along with Business By Design, but clearly they also see they need to be developing the expertise.
For the time being, there is still a tremendous advantage for newer players. It is more of an architectural advantage. I’m talking about both their software and their organizational architectures. As was mentioned on the program, it’s easier to start clean sheet for the Cloud than convert after the fact. The bigger the company, the harder to convert, and the slower that conversion must be.
“McKinsey recently said the Cloud is more style than substance because:
Nobody agrees on the definition
It doesn’t scale to Enterprise
It distracts attention from areas where tangible value can be unleashed.
Why would they say that? Are they being influenced because they’re in line with the old model and vendors?”
First, I did not think the McKinsey report reflected a very deep analysis. The coverage I saw on it was universally negative. From my perspective, they picked a conclusion and then drew up an analysis that supported their conclusion, so yes, I’d say they’re part of the Old School “Military Industrial Complex” around IT. They have an agenda.
SaaS eliminates a lot of value from the ecosystem for third parties like McKinsey precisely because it is service and that’s what McKinsey and the SI’s are in the business of delivering.
That particular report did a lot of silly things in analyzing the cost of the Cloud for larger organizations. The per-server cost for corporate IT were ridiculously low compared to many other estimates I’ve seen (just the power costs alone from data center studies I had seen were a big fraction of what McKinsey claimed). They burdened the analysis with a lot of costs that were irrelevant to the choice of Cloud or Data Center. That just added a lot of fixed costs that masked variable cost differences.
Given their great name, the study really doesn’t reflect very well on their expertise. But it will be a handy piece of collateral for those that want some air cover from the advantages the Cloud is bringing and the attendant disruptions that entails to an industry.