The first keynote presentation on Thursday was delivered by Tier 1 Research VP Dan Golding. As an analyst, he set out to give an overview of market conditions that might affect hosting providers, and in particular, he set out to address the cloud computing issue.

Rather than the fluffy angle (I’m not sure if that was a pun), he said, he wanted to discuss how cloud computing is affecting hosting providers right now, and how it’s going to be affecting hosting providers in the future.

He offered a definition of a “cloud” offering. Pretty standard stuff we’ve all heard before, but a couple of his key requirements were that it’s available on-demand (especially in terms of provisioning) and it has to have some kind of granular billing, along with a few other criteria.

Interesting point – he says you shouldn’t confuse enabling technologies with the product in the cloud computing business. For instance, a product that uses virtualization isn’t necessarily a cloud product, and a cloud product doesn’t necessarily use virtualization (though it probably does). Virtualization is just a technology, in this case, not the product.

He identifies “platform as a service” and “infrastructure as a service” as being separate categories of cloud computing. Infrastructure as a service is more akin to how hosts envision a cloud product – a replacement for specific infrastructure bits and pieces like servers. And the customers of those IaaS offerings will be the PaaS operators (his example of that kind of services is Facebook – a kind of platform).

As far as smaller hosts competing in the cloud space, he says there’s plenty of opportunity, because cloud computing is absolutely not a mature technology.

He identifies the big players, with the caveat that history has shown that while there is probably a significant first mover advantage, that doesn’t necessarily translate to winning in the space.

One of the things he really focuses on here is the missteps among the big players (Amazon, Google, Microsoft and Salesforce, by the way), all of which could mean opportunities for hosts to win.

The thing that’s interesting here is that a lot of these companies are businesses that posses massive infrastructure, and are trying to turn some of their capacity into a product. But they’re not necessarily built precisely to the specs of what cloud computing could or should or will be.

One of the key considerations is who the target markets for these services are. In Golding’s opinion, they’re SMBs, Enterprises, ISV and SaaS.

The other big question (and one that is really overlooked, he says, by a lot of breathless tech press coverage) is – do people really want cloud computing? Is it the best fit for everyone?

Tier 1 did some “unscientific polling” last year, and their results show that among hosting customers, almost nobody is currently using cloud computing. And there is no runaway leader at this point. The positive outlook would be that the market is really untapped at this point. The converse might be that there is actually no market for cloud computing (that was kind of a joke, I think).

He says mid-tier enterprises are the “sweet spot” for cloud computing (as customers). They have big needs and small resources for capital expenditures. They’re also going to be the big customers for SaaS applications.

Among applications, he feels the top SaaS applications (HR and accounting, messaging and collaboration, CRM, project management, etc.) are the things that are definitely going to move to the cloud, as a platform, and are the subjects best suited to them.

In discussing what hosting providers can deliver in a cloud computing strategy, he showed a pretty interesting chart, outlining what features were important to hosting customers. By far the most important elements were price, flexibility and speed of provisioning. A few of the things that didn’t register at all were brand and “green IT.”

Here is one of the things he really wanted to emphasize: a look at who, among managed hosts, is doing a good job, or an interesting job, in the cloud computing space. They included OpSource, Rackspace, StrataScale, Flexiscale, Terremark, Savvis, Layered Tech, Mezeo and Nirvanix.

One of the first keys to winning in the cloud space is making sure you’re doing all the early maintenance related to security, licensing and other issues that could be sources of failure. Other keys, according to Golding, are identifying a good target market, and riding the new wave of virtualization technology (as enabling technologies go, virtualization is pretty key).