Cisco Gets Serious About IaaS

Given all the things that I'm engaged with these days, it was rewarding to see our friends at Cisco make a good set of announcements around enabling service providers with IaaS -- Infrastructure as a Service.  And I'm a bit embarrassed that I didn't catch this earlier ...

But -- at the same time -- as I digest this announcement, I'm torn.

There's a lot to like here.  But it also reinforces my notion that we -- as an industry -- have just started to travel down a rather long road ...

The Announcement

Perhaps the best overview is this post from Simon Aspinall that tells the story.  Basically, it's the Cisco  products that are relevant, plus some interesting services as a wrapper.

In particular, I found this online calculator particularly interesting -- it's a handy presentation of the underlying cost drivers, and certainly a useful thing.

What's Good About This

Clearly, Cisco (as well as EMC and other vendors) have figured out that service providers -- of all sizes and stripes -- will be increasingly important in the future.

Some of their buying criteria will look familiar to enterprise IT types.  And certain things will end up being very different.  And it's nice to see Cisco start to address the differences, rather than similarities.  I think many of us infrastructure vendors are starting to do the same thing.

The offerings and solutions are very network-centric (as they should be), reflecting Cisco's strong capabilities in this domain.  And one could create an argument that getting the network model and technology right is a must-have for any service provider of decent scale.

All in all, a very promising start from a strategic vendor in a burgeoning market.

But There's So Much More To Do!

I guess I've been spending so much time thinking about this space -- as well as talking to service providers -- that my perspective has become, well, rather focused!

So many of us in the IT infrastructure business are focused on the "supply" side of the equation, e.g. what are the underlying costs associated with providing a given service?

But what service providers are really interested in is the "demand" side of the equation: what services are being demanded by the market, how much will people pay for them, how will they get sold, how do you reach customers, etc. etc.

It's a good thing that Cisco has started to make at least a few of the ingredients in an IaaS-based model more approachable.  That's goodness.  Needless to say, you'll need a few more ingredients to make a complete IaaS "meal", but it's a start.

Me?  I'm spending a lot of time creating an array of business model sequences that assume a starting point (e.g. outsourcer, telco, reseller, pure-play, vertical, etc.) and an implied set of entry-expansion-maturation sequences.  Given that I spend a ton of time with classic enterprise IT, I've got a good sense of what people might want, and what they won't.

Lots of complex variables in play as I struggle with this, including the unavoidable uncertainty in attempting to predict the future.  My head hurts, but I'm starting to make some good progress.

Putting The Pieces Together

I think that the winners in this IaaS SP space will be able to do two things well: show an aggressive "cost-to-serve" advantage for their piece parts, as well as set of differentiated capabilities that translate into progressively more differentiated (and presumably profitable) implied business models.

Cisco (as well as EMC, VMware and other vendors) have started to address the first part of this equation.  The fun part will be figuring out the second part.

We live in interesting times.