Getting real on colocation
Of late, I’ve had a lot of people ask me why my near-term forecast for the colocation market in the United States is so much lower (in many cases, half the growth rate) when compared with those produced by competing analyst firms, Wall Street, and so forth.
Without giving too much information (as you’ll recall, Gartner likes its bloggers to preserve client value by not delving too far into details for things like this), the answer to that comes down to:
- Gartner’s integrated forecasting approach
- Direct insight into end-user buying behavior
- Tracking the entire market, not just the traditional “hot” colo markets
I’ve got the advantage of the fact that Gartner producing forecasts for essentially the full range of IT-related “stuff”. If I’ve got a data center, I’ve got to fill it with stuff. It needs servers, network equipment, and storage, and those things need semiconductors as their components. It’s got to have network connectivity (and that means carrier network equipment for service providers, as well as equipment on the terminating end). It’s got to have software running on those servers. Stuff is a decent proxy for overall data center growth. If people aren’t buying a lot of stuff, their data center footprint isn’t growing. And when they’re buying stuff, it’s important to know if it’s replacing other stuff (freeing up power and space), or if it’s new stuff that’s going to drive footprint or power growth.
Collectively, analysts at Gartner take over a quarter-million client inquiries a year, an awful of lot of them related to purchasing decisions of one sort or another. We also do direct primary research in the form of surveys. So when we forecast, we’re not just listening to vendors tell us what they think their demand is; we’re also judging demand from the end-user (buyer) side. My colleagues and I, who collectively cover data center construction, renovation, leasing, and colocation (as well as things like hosting and data center outsourcing), have a pretty good picture of what our clientele are thinking about when it comes to procuring data center space, in addition to the degree to which end-user thinking informs our forecast for the stuff that goes into data centers.
Because of our client base, which not only include IT buyers dispersed throughout the world, but a lot of vendors and investors, we watch not just the key colocation markets where folks like Equinix have set up shop, but everywhere anyone does colo, which is getting to be an awful lot of places. If you’re judging the data center market by what’s happening in Equinix Cities or even Savvis Cities, you’re missing a lot.
If I’m going to believe in gigantic growth rates in colocation, I have to believe that one or more of the following things is true:
- IT stuff is growing very quickly, driving space and/or power needs
- Substantially more companies are choosing colo over building or leasing
- Prices are escalating rapidly
- Renewals will be at substantially higher prices than the original contracts
I don’t think, in the general case, that these things are true. (There are places where they can be true, such as with dot-com growth, specific markets where space is tight, and so on.) They’re sufficiently true to drive a colo growth rate that is substantially higher than the general “stuff that goes into data centers” growth rate, but not enough to drive the stratospheric growth rates that other analysts have been talking about.
Note, though, that this is market growth rate. Individual companies may have growth rates far in excess or far below that of the market.
I could be wrong, but pessimism plus the comprehensive approach to forecasting has served me well in the past. I came through the dot-com boom-and-bust with forecasts that turned out to be pretty much on the money, despite the fact that every other analyst firm on the planet was predicting rates of growth enormously higher than mine.
(Also, to my retroactive amusement: Back then, I estimated revenue figures for WorldCom that were a fraction of what they reported, due to my simple inability to make sense of their reported numbers. If you push network traffic, you need carrier equipment, as do the traffic recipients. And traffic goes to desktops and servers, which can be counted, and you can arrive at reasonable estimates of how much bandwidth each uses. And so on. Everything has to add up to a coherent picture, and it simply didn’t. It didn’t help that the folks at WorldCom couldn’t explain the logical discrepancies, either. It just took a lot of years to find out why.)
Posted on April 23, 2010, in Infrastructure and tagged colocation, Gartner, resesarch. Bookmark the permalink. 14 Comments.
CloudPundit: Massive-Scale Computing 
Lydia,
Intruiging post. I see that there is a tremendous capacity strain in many mature markets, and virtually no investment in ’emerging’ or ‘undiscovered’ markets. You selected 4 areas in particular that I wanted to delve into further:
IT stuff is growing very quickly, driving space and/or power needs
My observation here is that IT stuff is shrinking quickly in footprint and conversely growing in terms of power draw. Blades are becoming more and more utilized by organizations looking to maximize density and as a result they pull more power than a 1 U pizza box. They are also WAY heavier, so floor loads become an issue quickly as well.
Substantially more companies are choosing colo over building or leasing
I see heavy activity in both areas. Single tenant facilities aka ‘campuses’ are being built all over the country for their corporate parent and while this is good for that company, it will not trickle down to broad based metrics and hot colo markets because they only (need to) care about one customer – themselves. Where the hot markets are is where there are multi tenant facilities – lots of companies can lease a chunk of a 200k foot building. Those areas are tight, and the tightest segment I believe are companies who need 6-20 megawatts, there are as few as 6 potential sites in the US to support them right now.
Prices are escalating rapidly
The company who has the inventory drives price. Factor in how long the inventory has been on the market, the other deals in the market that were done over the past 1-2 quarters, what the terms were of the financing for the inventory – those all play roles in the overall price. Someone who has a balloon payment due who needs a deal to trigger a refinance is going to take a lower price than someone who has 1Mw left and is throwing off a ton of cash.
Renewals will be at substantially higher prices than the original contracts
I am starting to see this happen and the most prominent variable was when a company signed their initial lease. If they did a 10 year lease with 3% bumps then the sticker shock may not be as bad as someone who did a 2 year lease in a building that is almost full because that space has become more valuable.
Other things not mentioned that are becoming popular and maybe indicative of tightening markets is the number of ROFO/ROFR requests seen in the ‘hot markets’ and the fact that roughly one third of current colocation leases come up for renewal in the next 18-36 months. People know the space will be tight for a while a need insurance against having to expand and no place to do it so they are asking for Rights of First Offer and Right of First Refusal in most deals. When the leases start coming up – depending on how much inventory your provider has – be prepared to pay more to be the incumbent, or be ready to explore a move. if you’re in a hot market you’ll be OK, ’emerging’ market. Not so much.
One last thing – there are good reasons that a lot of space exists in the non-hot markets – hard to get to, no skilled talent pools, remoteness of power, connectivity or both. I know of two old paper mills in Maine where electricity is half a cent a kilowatt hour. Abundant, cheap green hydro power, but no network and no full time residents within 100 miles. Literally. I think most companies would prefer being in Northern Virginia and why that will be a hot market for a while.
LikeLike
Interesting article, and I think as new companies spin up, they will use EC2 or various spin off projects to either get off the ground and/or to grow. Until they hit a point where they need their own data center space of significant size.
Also, I think companies will figure out ways to “virtualize” some of their needs in order to save the power/space footprint.
LikeLike
On your points, Mark:
1. Yes, absolutely, power and not space is becoming the primary limiting factor, especially as virtualization shrinks footprints. Floor loads are generally secondary to this, since in many cases if this is my issue in an enterprise data center, I have enough space to spread my gear, and I’ll probably run out of power first. There’s substantial net-new space needs that are being driven by “the data center I have is obsolete and I need to replace it”, but most of the customers with that problem go into a lease or build arrangement, rather than colo.
2. Leasing has become very, very hot over the last 18 months. There still aren’t enough high-density data centers available (either in the colo or lease markets), but it’s a very modest percentage of overall demand.
3. Colocation pricing in general has been, and is likely to remain, volatile, and the smaller the market the more volatile it gets. The financing you cite plays into it. So does where someone is in their fill; we continue to see pretty aggressive opening specials, for instance, which can temporarily depress prices in an entire metro market.
4. Leasing prices have lots of room to move upwards because of the longer term cycle of leases. Colo contracts are usually only 2 to 3 years; the trend towards 5 years is very recent. So 2010 and and 2011 renewals will mainly be for deals signed in the 2007-2009 timeframe, and pricing has bounced around a fair amount but isn’t, in most places, wildly above 2007 levels. When we see straightforward same-provider renewals right now, it’s often at right around the same price point as the previous contract; if the customer is willing to move, they can often get lower prices within the same city.
I agree that ROFO/ROFR is a factor in hot markets, and can be a factor in other markets as well, depending on where the data center is in its fill.
The big growth trend we see these days is in the lease and the wholesale side of the market, rather than retail colocation. What is capping the leasing and wholesale side is supply, rather than demand. I have a ton of clients who would really like turnkey space in their local cities (because they usually plan to seat their IT staff alongside the data center), or who would like it within closer commute distance of their offices in a major metro market.
Remember that the non-hot markets are not necessarily rural; Maine is a poor example in that regard. They’re simply second and third-tier cities, often “NFL cities” with a solid population of businesses.
LikeLike
It’s interesting that you cite the wholesale market – I am curious what that means. 1+Mw/10k ft or 5+Mw, 50,000 ft? In any case I agree that in what I would consider the second tier cities – Cincinnati, Minneapolis, and Detroit, there is little space of multi megawatt size. A half cabinet, no problem. Half an acre, no way…
The other thing I continue to scratch my head on is the ‘commute distance’ point. Unless you are a data center operator there arent many reasons to set foot in a facility maybe 3 times a year. The lack of office space, the time spent there is typically concentrated, and the fact that once the gear is in, a remote hands reboot should be about it. I would be curious to see how many of those companies who want to watch their lights blink have utilized remote management tools and associated best practices to the fullest.
Cloud starts to introduce obsalescence up the stack at the app layer so that will be fun to watch, although to me Cloud is just an accounting exercise.
LikeLike
We generally consider wholesale to begin at 5,000 sqft and up, though we’ve seen some wholesale-style deals in as little as 2,000. We use the term “large-footprint” for these deals — generally larger, longer terms, fewer amenities.
A shocking number of people are server huggers. I always ask our clients, “How often does your staff go into the data center?” If the answer is anything more than “very rarely”, I ask, “What are they doing there?” If people tell me they go touch their gear daily, or worse still, multiple times daily, I *really* want to know what they’re doing, especially if they also tell me that their infrastructure is mostly virtualized. Much of the time, IT management has no clue what their staff is doing in there, but they’re insistent that they’ve got to touch that stuff and so need to be close. (I mean, I encounter clients where I say, you’re virtualized, you’re not adding capacity daily, you’re not augmenting your existing hardware with more RAM or more drives or whatever, you have remote power management, what could your staff possibly be *doing* in there?)
LikeLike
I think that the IT infrastructure has gotten stale in their feeling they need “data center space” and that many folks who visit the data centers are doing very routine tasks and aren’t being very good in the time management field.
Watching some of the new startups basically spin up with some code and a bunch of EC2 instances, they can scale to the point when they need datacenter space, they go pretty large (look at twitter as an example).
Mark, many IT management folks need to take people on tours of their data center. If they are a startup, Investor types, VCs, other companies, etc. If they are established, it’s to show their management they have real stuff. For some reason, having a data center in Kansas is too scary for folks.
LikeLike
The entire colocation industry still astonishes me. Unless customers are smaller users of space, and need a TIER 3 or 4 center which could be cost prohibited, are limited with time to build a new space, are in need of temporary expansion space, or are start-ups and unclear of their long-term needs. What’s driving companies to move toward colocation centers in masses?
I believe there are some misconceptions reported regarding the cost of building a new data center. Mostly, I read about costs ranging from $1,000 to $2,000 per sf. I’m not sure where these numbers are coming from (could they be from the colocation folks themselves)? In practice I’ve seen TIER 3 rooms as small as 5,000 sf at 150 W/sf for under $600/sf. There’s no doubt that higher heat loads are more costly, but unlikely to the extent that makes renting less costly. After all, someone is making money or they wouldn’t be in the business. Certainly, companies look at cost of money, and having manpower to manage them, however, with the high cost of colocation which will only increase, will large users (5,000+ sf) really be hoodwinked into long term large spaces.
It would be great to see a true, third party analysis of the true apples-to-apples cost to own and operate a data center v. lease…recognizing this varies by region based on real estate cost, this too could be factored in, and showing results for a variety of cities. Perhaps if the real costs were exposed there would be a significant effect on the expansion of the industry.
LikeLike
Think you started out with a very interesting very strong statement then fizzled. I could care less about your WorldCom predictions. A white paper or report supporting your colo assertions would be invaluable.
Cheers,
LikeLike
This design is steller! You most certainly know how to keep a reader amused.
Between your wit and your videos, I was almost moved to start my own blog (well,
almost…HaHa!) Excellent job. I really enjoyed what you had to say, and more than that,
how you presented it. Too cool!
LikeLike
Actually when someone doesn’t be aware of then its up to other people that they will assist, so here it takes place.
LikeLike
Pingback: Internet Marketing Income Tools » The Magical Power Of SEO Link Vine
Pingback: Gartner: Getting real on colocation | Lifeline Data Centers
Pingback: Getting real on colocation « CloudPundit: Massive-Scale Computing | Wordpress Blog Hosting
Pingback: Gartner: Getting real on colocation | Lifeline Data Centers -