10 Cloud Secrets For IT Service Providers

As part of my role here at EMC, I get to talk to all sorts of interesting consumers of IT: enterprises, partners -- and, more recently, the more nascent category of "IT service providers".

I think of service providers as a new channel for the IT industry.  Yes, we've seen outsourcers and hosters and all manner of people who will offer to run various aspects of IT on your behalf, but enough has changed recently that the whole discussion is somewhat fresh again.

Given that I spend a lot of time with enterprises and more traditional partners as well, I have a pretty good sense of what people might want, and why they might want it.

And, in this post, I'm going to share what usually comes up in the first hour of discussion.

Agree or disagree, as you see fit!

#1 -- It's A Business Discussion, Folks

So much of the "IT as a service" SP discussion quickly devolves into this technology, or that technology.  Sure, there are clear technical requirements for these use cases, but -- ultimately - SPs are in this to make a buck or two.

The implication is clear -- the technology agenda is driven by the business agenda, and not the other way around.  It sounds obvious when you say it, but I find myself saying it more than I should be :-)

#2 -- It's All About Differentiation

So, you want to offer a cost advantage by building a bigger / cheaper / better data center?  The line forms over here, please. 

Frankly, some of these models will work because of happenstance (you happen to own gobs of dark fibre, or you somehow acquired empty data centers near hydro power, or similar), but a lot of them won't be defensible in the long term.

At the end of the day, everyone pretty much pays the same for the core inputs.  Cleverness in assembling the bits isn't defensible intellectual property.  Sheer size isn't a defensible position, either.  Anyone who wants to play that game can do so -- given enough capital.

Besides, if you spend any time with enterprise IT types, cheap isn't the only thing driving their agenda. 

Nope, in my humble estimation, you've got to do things that (a) the customer can't do or doesn't want to do, and (b) do things that your competitors can't easily follow you. 

Call that "differentiation" -- that's where you make money.

IT vendors find it hard to make a living when they sell undifferentiated commodities.  IT service providers will undoubtedly find the same thing to be true.

#3 -- Where Will The (Paying) Customers Come From?

I can't tell you how many times I've had a "field of dreams" conversation, e.g. build it and they will come.  Well, there's more to it than that.

Most SP business models are about delivering efficient IT services at scale.  This means that there's not a lot of money in the business plan for high-touch customer engagements that most IT organizations have come to expect -- especially when trying something new or different!

Indeed, this go-to-market aspect quickly becomes the central discussion when I meet with SPs.

#4 -- How Will You Put Customers In Control?

One thing that differentiates enterprise IT from other forms is control: service delivery, resource consumption, cost containment, security, compliance, etc.  Most IT organizations are willing to have others do certain aspects of the job ; the one thing they are most reluctant to give up on is control.

Indeed, part of my rant on a recent discussion around "secure multi-tenancy" focused on this very fundamental point.

It's a logical perspective, and not unique to enterprise IT -- you'll find the same sort of thinking in other enterprise function: marketing, legal, manufacturing, customer service, etc. etc.

Getting this "customer is in control" thing is the key difference in getting a decent shot at workloads and services that matter, and ones that don't.

#5 -- How Will You Cross-Sell and Upsell?

The most expensive part of most IT-related business models is customer acquisition.  So, you've got to put serious thought into the sequence of offers that help you from an initial customer to a fully engaged customer.

Put differently, it's not enough to have a bunch of unrelated offers for different customer segments.  Ideally, you'd have a sequence of offers that target the same customer segment, but provide increasing (and related) value.

And that takes a bit of thought, and a bit of insight :-)

#6 -- Make Your Offer Convenient To Consume

At the polar extremes, we've got traditional outsourcing (inherently complex and difficult to consume) and self-service compute models (inherently easy to consume). 

Somewhere in between there's an interesting landscape of more significant IT services that respect the easy-to-consume model.  And there is where I tend to think all the attractive opportunity will lie.

Why make your offering convenient to consume?  Because things that are convenient to consume tend to get consumed more!

#7 -- Focus On Your Secret Sauce

Lots of people can build and run IT infrastructure.  Lots of people can package up various IT services and make them ready to consume.  Just about any aspect of an IT-as-a-service model can be out-tasked to others (and often is), if you think about it.

Ask the hard question -- what's the secret sauce in your business model today?  In a year?  In five years?  What parts do you want to own, and what parts might be better left to others to go do. 

Before I meet with a service provider, I can usually figure out what a good answer might be -- but it's not enough for me to think I know -- you should know as well!

#8 -- Plan For The Future

I don't want to mention names, but I've seen a few ugly scenarios where one SP business idea after another is tried, each resulting in a wildly incompatible stack of technology and operational models.  You wouldn't want to build enterprise IT infrastructure that way, and you wouldn't want to build IT-as-a-service infrastructure that way, either.

Each technological / operational decision should ideally be a "two-fer": meet the needs of the immediate opportunity, as well as lay the foundation for opportunities and offerings to come. 

Besides, that's how many enterprise IT organizations think about it.

#9 -- Partner With Your Vendors

IT vendors (especially the good ones) are good for more than technology bits and pieces. 

For one thing, we all work with enterprise IT users, so we sort of know what they want.  Our brands and capabilities are known to them, and often the exact same stuff they're already using internally.  And we're pretty good at identifying the propositions that they'll consider, and those that they'll flatly reject for all the right reasons.

Sure, we want to sell you a bunch of cool stuff.  But the more serious vendors will recognize that we won't be successful unless the service provider is successful -- just like any other channel partner!

#10 -- Move Fast, Invest Wisely, And Prepare To Learn A Lot

Committing big piles of money to a single business idea isn't necessarily a good thing if you're one of these service providers considering different forms of IT-as-a-service.  The market is moving fast, and customer perceptions are changing quickly.  The payoff could be good, but also is the associated business risk.

Better to get into the market quickly with a modest investment, watch what happens, and "fail fast".

Final Thoughts

So, there you have it -- my sincere advice for anyone considering an IT-as-a-service offering.  There's more to it than this, naturally, but these are usually the thoughts that have to be shared before we get into the fun discussion -- creating and offering new services that customers say they want.

Comments

[nate]

Cloud secrets for customers

- Don't expect any provider out there to be able to provide you with an "infinite" amount of on-demand capacity and pay pennies on the dollar for.

I just got off yet another call with another cloud company where management here has some stupid expectation that SOMEHOW, SOMEWHERE there is a cloud company out there that can do what we need cheaper than we can do it ourselves.

EVERY SINGLE ONE IS THE SAME, all of them are 2-5x and even 10x more expensive then doing it in house. And after all of the cost analysis I have seen done and conference calls and presentations etc, if anyone came in at/below our cost I would call bullshit on them(over selling what they have no prayer to be able to deliver). Even Amazon is way above our cost, with their crap shared infrastructure on poorly available servers.

Yet they(management) continue to persist on a quest for the holy grail.

IT DOESN'T #@$@ EXIST. Well maybe it does, but it's in the form of leveraging the resources you already have in house.

I suppose sometimes I take my job too seriously.

I need a drink.

[Chuck Hollis]

Nate

Your points are valid. Many IT organizations are of enough size and scale that -- yes -- they can be more cost-effective than external providers, should they choose to.

That's one of the tenants of the private cloud model, it doesn't presume that external service providers are inherently better.

I'll join you for that drink ...

-- Chuck

[Chuck Hollis]

All ...

Was browsing this morning and came across this horror story of a failed RAID card, corrupted data, backups that can't restore, etc.

I suppose I should add #11 -- "Be At Least As Good As The Enterprises You're Serving"

http://www.theregister.co.uk/2010/02/10/hostv_outage/

-- Chuck