I was in a briefing today, and a customer asked an interesting question.
She wanted to know why it seemed like the big server guys (IBM, Sun, HP) were only half-heartedly going after the storage market, and weren't investing in storage technology, services and support the same way that focused players (e.g. EMC and others) were.
We ended up talking about different aspects of the industry, and I guess they found it pretty interesting, so I thought I'd share it here.
Interest vs. Investment
I think it's fair to say that the server guys are interested in storage in a big way. They see the money that gets spent on storage, backup, replication, management, archiving, etc. and it's a pretty big number. And they don't like competing vendors in their accounts.
But there's a difference between "interest" and "investment".
Investment means you invest in R+D to build your own products, rather than reselling others.
Investment means that your sales and pre-sales people are focused on storage across a variety of server environments, and not just the ones that are being sold.
Investment means a focused qualification and support capability that's very storage literate and server agnostic.
These are big investments for any company to make, and -- for those of us that are close to the industry -- it appears that, generally speaking, the storage-focused vendors take these sorts of investments more seriously than server companies in general.
Now, in all fairness, there seems to be pretty substantial investment in the storage marketing side of the equation for these server companies. But if you look behind the scenes at the actual products, technologies and services coming out, and the level of qualification and support, and the solution building capability -- marketing alone can't hide the general trend.
So why is that?
Market Reach
If you're going to invest in designing, building, selling and supporting storage solutions, you're going to want to amortize that investment as much as possible.
But that brings up a thorny problem when you come to distribution. If you're a server vendor, most of your effort goes into selling your servers and the stuff that goes with them. Your reps, channels and partners are all set up around selling a server-centric solution. Storage is an interesting sideline.
You're not going to see an IBM, Sun or HP rep going off-base very often to sell their storage products behind a competing server platform. At least, not more than once ...
So, from a business model perspective, you're stuck. You either make a massive investment in a dedicated sales and support organization to go off-base, or you reduce your storage product investment to comfortably fit within the paramaters of your server business.
Which means that the big server vendors end up heavily relying on reselling products that other people build, and limiting their investments on products that they build themselves.
OK, you might see a product here, or a "first" there -- but it's a pretty broad landscape that needs to be addressed. And if you look at the patterns, rather than the press releases, a different picture emerges.
Competition on R+D Spend -- And Talent
At every company, there's a bucket for R+D, and different interests compete for where that's spent. At a server company, if it comes down to making sure that the servers are great vs. making sure that the storage is great -- well, I can't make specific predictions, but I can guess as to how those wrestling matches work out.
Fall behind in the incredibly competitive and aggressive server market, bad things happen that get a lot of management attention. Fall behind in the storage investment; well -- lots of storage gets sold packaged with servers, so it's under a bit less pressure. And no company can invest in everything.
Another problem comes with talent. If you think you're a hot-shot storage engineer, where do you want to work? Or if your a great storage presales or support guy? Sure, all of these server companies are good places to work, but if you've decided that you're a storage professional, all things being equal, what's your preference?
Occasionally, we'll see IBM, Sun or HP go through an internal re-org, and the storage product guys end up working for the server product guys. Or maybe the same thing will happen in the field. From a purely selfish perspective, we like that sort of stuff, because we know what the inevitable outcome will be.
Best Of Breed vs. Packaged Solution
Since I've been in this business, I've noticed that there's a part of the market that just goes ahead and buys whatever storage the server vendor offers. Maybe they see this as getting a better deal, or support is easier, or they feel loyal, or the integration will be better, or something along those lines.
Nothing wrong with that -- I get it.
And, trust me, a whole lot of the storage that IBM, HP and Sun sells falls along these lines. Good for them.
But I would argue that storage is becoming more and more strategic going forward, and more and more customers would like to treat it as an independent decision.
At one time, server vendors built and sold the majority of operating systems. Not anymore.
Ditto for networks. And applications. And databases. And management tools. And so on.
All of these were once businesses where the server vendors were dominant.
And all of these technologies shifted into a best-of-breed model over time. Storage seems to be marching down the same path.
The inevitable result is that, if you're a server vendor, your (remaining) storage customers are not that much interested in things like new features and differentiation (those people are already talking to the specialists), but tend to be more interested in things like convenient packaging, convenient purchasing, convenient support, integration with their favorite server, and so on.
The storage offerings from the server vendor seem to move downmarket and become more commodity and appliance-like over time, because that's where they're selling.
The Bottom Line
Every company on the face of this earth is in business to make money. And every company has to figure out where it needs to invest to do this.
Server vendors invest heavily in server technology, and over the long term invest less so in other areas, including storage.
I thinkl from a business model perspective, it's inevitable.
I think server vendors think storage is very important. And they want to make a lot of money selling it.
But that doesn't mean they're investing in it.
Now, in all fairness, not everyone needs to buy from a storage specialist. There are lots and lots and lots of customers whose needs are served by a combined server/storage package from their server vendor.
But I think the trend is inevitable, as environments get larger, and more money gets spent on storing and managing information, that specialists will do better in the long run than generalists.
What do you think?
You mentioned big server vendors HP, IBM and Sun, but you neglected to mention one very important one. How about Dell - they are the #1 vendor for industry standard servers in the USA and EMC has a very strong OEM and reseller relationship with them. Where do you feel they fit in your thoughts?
Posted by: Geoff Mitchell | July 31, 2007 at 06:14 PM
Ahhh .. I thought that question would be coming.
Yes, we work very well with Dell, and FSC, and Unisys, and NEC, and a few others.
I know this is going to sound a little soapy, but -- in each of these cases -- I think that the products are clearly identified as EMC products, and much of the extended EMC value proposition comes through: qual, support, consulting, services, integration, etc.
Simply put, the customer knows what they're buying.
I tend to think of this sort of relationship as different than a standard OEM arrangement where the server company tries to convince the market that the product is their own through naming or branding (insert lots of examples here).
That being said, I think that all of these relationships are win-win, in that each company benefits, and -- ultimately -- the customer does as well.
Thoughts?
Posted by: Chuck Hollis | July 31, 2007 at 07:00 PM
Chuck,
I couldn't agree more on this one (evidence here - http://blogs.rupturedmonkey.com/?p=77) ;-)
As storage gets more and more important I see an increasing market for best of breed/specialists.
Game on.
Posted by: Nigel Poulton | August 01, 2007 at 01:32 PM
Game on!
Posted by: Chuck Hollis | August 01, 2007 at 02:17 PM
Large scale storage is always going to be better bet from a specialist like EMC. But an interesting twister coming your way is VmWare running on massive servers with direct attached storage in share-nothing clusters.
Posted by: Fred San | August 02, 2007 at 12:46 AM
Hi Fred-san:
Share-nothing clusters have been around in one form or another for more than a decade, and if VMware decides to go there, it'll be yet another variant.
The computing problems they tackle well are interesting, but rather limited in scope from a market perspective, and -- up to now -- haven't made much of a dent in the storage market.
I posted on this a while back (see "Seduced By Scale Out").
Every year it seems some part of the IT landscape gets seduced by the idea some flavor of scale out. Some people think it's gonna change everything.
And, so far, it hasn't ...
Posted by: Chuck Hollis | August 02, 2007 at 07:53 AM