Beyond Building Count?

September 26th, 2011 by · 27 Comments

On Thursday, Level 3's CEO Jim Crowe was asked about on-net building count (~8200), and he took the opportunity to comment the metric itself and offered an alternative:

Personally I think you should tell people how many square feet you have on that as opposed to buildings, because two or three buildings here in New York are worth 50 buildings in Peoria. That is a much, much more relevant metric: how many square feet of enterprise building.

A very good point!  While square footage on-net isn't perfect either, it would be a very valuable metric in understanding fiber-based business models.  The only reason I collect building count as a starting point is because the data is there, and it does at least give us a sense of who is digging up the streets and who isn't.  But I'd collect square footage on-net in a heartbeat if the data were available.  Then I'd take a look at metrics like on-net square footage per route mile, per fiber mile, etc to see what's in there - and I have only a vague idea what I'd find.

However even Level 3 doesn't offer up that data yet that I know of. The only place I recall hearing similar data offered is from Cogent's Dave Schaeffer at some presentation in the past that I'll try to dig up.  But that doesn't mean we can't try to start a new trend!  So I'm going to take Jim Crowe up on his suggestion.

I hereby request the square footage of enterprise buildings on-net from any fiber operator willing to share it -- public or private, ILEC or CLEC or MSO, national or local, foreign or domestic.  Any takers? Comments for or against?

If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!

Categories: Metro fiber

Join the Discussion!

27 Comments So Far


  • GigabitG says:

    Sounds interesting, but I don’t really understand how carriers are going to be able to measure the amount of square feet per enterprise building, or whether it really matters that much anyway.

    Far be it for me to argue with Mr Crowe, but if the building is on-net, it’s on-net, and it doesn’t really matter if the carrier just has a fibre up the risers, or a fully fledged suite within the building. It’s on-net and for all intents and purposes, that carrier should be able to offer lit services quickly and cost-effectively.

    How about devising a metric around the type of enterprise customer per on-net building. Eg, if a carrier says it has 100 financial enterprises on-net, that’s quite different from saying it has 100 residential apartment blocks on-net.

    • Rob Powell says:

      Actually, I don’t think Crowe was suggesting just the square footage of on-net enterprises served, but the whole building once it’s on-net for any one enterprise within it. That represents the on-net opportunity already in place, and would be what we’d want to track.

      For the ‘type of enterprise customer per building’, that would get messy for multi-tenant locations I’d think. Not that I wouldn’t love to have that sort of data too…

  • Rob Powell says:

    Raw building count numbers do tend to overstate the relative size of operators that focus on smaller markets. Level 3’s building count is on both sides of that fence though. They have a substantial tier 1 presence, but they also have lots of on-net buildings in places like Winchester VA from the Telcove assets.

    • en_ron_hubbard says:

      This has been an obvious weakness in reporting for a long time — basic industry definitions should ideally be synced so that comparisons are meaningful– and I’m glad that Crowe agrees.

      Being connected to a 100,000 ft2 multi-tenant OB offers far more potential than being connected to Moe’s Oil & Lube building in Yougstown.

      Towers should also be reported as a separate category.

      Cogent does do more detailed reporting– giving, for example, average sq footage for their reported lit office buildings. The most informative stat would be sq footage served as a percentage of sq footage available per location. This would show the cheap (to connect) revenue within the lit geography.

      If LVLT initiated this type of reporting perhaps others would be forced to follow.

  • Raul Martynek says:

    Rob, this reminds me of the old days when we raised money as a BLEC (Building Local Exchange Carrier) and used these types of metrics. If you really want to get detailed about it you would not only capture sq.ft. but total tenants in the building and total tenants that are customers. We had about 450 buildings “lit” and used to capture all that info. It’s not too hard, plenty of real estate guides/sites that know sq.ft., owner, and number of tenants.

  • FBLDan says:

    I would concur that square footage would also need to include tenant numbers to be realtive. A single tenant occupying 100,000 square feet is much different than a 100,000 square foot building housing 10 tenants.

    • Rob Powell says:

      I suppose the problem here is finding metrics that are both helpful to our understanding, but is also easy to collect across the industry with a reasonable expectation of similar assumptions underlying it. Are you suggesting we collect three numbers: 1.building count, 2. average square feet per building, and 3.average tenants per building?

  • Kevin Rocks says:

    Perhaps you might consider renaming the lists “fiber asset list”, or “bldg square footage list, because that data does not reflect a carriers ability to DELIVER lit services. “The “lit” building list is perhaps one of the most inaccurate and overstated “metrics” in the business. What lit means to one provider, does not mean the same to another. A free standing building in Decatur, Illinois is not the equivalent of a Class A High Rise in a Tier 1 city. Also, just because a carrier has fiber in the building does not mean they can service each and every floor, far from it. In New York City for example, every landlord has rules and regs in terms of riser fiber, core drilling, etc. In the carrier hotels, it is even more pronounced. At 111 8th, just because a carrier is present on the 3rd Floor, doesn’t mean they can also provide lit services to each suite even on that floor, within a reasonable installation interval, ie 60 calendar days. The most accurate metric is to list each building by floor and by suite, and confirmation on what services a carrier CAN DELIVER to those floors/suites within 60 calendar days. By default, this captures data regarding the EQUIPMENT the carrier has IN PLACE to deliver services within 60 days, a metric that no one seems to mention on this thread. If there is a 90 day lag to have the equipment in place to deliver a GigE or an STM-4, what good are fiber and square footage metrics?? Without this information, square footage, fiber miles, etc are essentially raw data listing a carriers assets, not their ability to deliver lit services. The only metrics worth stating in terms of “lit” buildings are what specific services a carrier can deliver, to which floor/suite in a building, and in what time specific frame.

    • en_ron_hubbard says:

      Kevin,

      You make good points, but even “baby step improvement” in reporting would be a great advance. See below for what Cogent has made available:

      <>

      Now, there is a lot of missing info within that blurb but if we had similar metrics for all competeitive providers we would have a far better feel for the general “quality and density” of respective providers and ON AVERAGE the respective growth potential and cost thererof.

      The point is that such info would give far more clarity than exists today– certainly not perfect but better. Agreed?

      • en_ron_hubbard says:

        The blurb was missing. Here it is

        “Over the next few years, CCOI plans to grow lit buildings to 1,300 (from 1,189) long haul route miles to 60K (from54K), and metro route miles to 17.5K (from 16K). A typical connected building has 41 stories, 51 tenants, costs$120K to build, and has a 15‐18 month payback. CCOI averages roughly nine customers per building (18% penetration), which includes lower penetration from recent building expansion. The network and building expansion should drive growth for several years, as the company has attained 18% penetration in existing corporate locations and it continues to cherry pick locations that have the highest growth opportunities.”

        • Rob Powell says:

          One could argue that Cogent is able to give that statistic because it focuses so tightly on a single type of enterprise building. A similar ‘typical’ building for a more diverse metro operator would have less meaning, yes?

          • en_ron_hubbard says:

            Perhaps true, but if we had the following metrics for all carriers we would be far better informed:

            — on net square footage by building type
            — revenues/square foot by building type

            Perhaps also, the fact that CCOI can provide that type of info is illustrative of the benefits of their business model? I dunno…just maybe?

    • Hammerdinger says:

      Kevin Rocks – Has a good handle on the “on-net” discussion, this is the best observation relating to on-net I’ve seen on TR so far. I would take it a step further & address 2 issues. 1st, carriers should analyze their addressable market in their on-net building portfolio by calculating revenue per square ft. Muliti-tenant class “A” commercial buildings have higher revenue per sqft potential than an industrial property.2nd ask the question/report, does the carrier have the right to sell & service other tenants in the building via a building access agreement or are they required to run home runs from the MDF to the tenant. I have found that very seldom does the CLEC spend the money to install a distribution system in a building like the ILEC enjoys. The purpose for on-net capital is to drive lower cost (install & ongoing), faster installs & better control of the customer experience, not to hang trophy’s on the wall for Wall Street. Most of the buildings reported as on-net in our industry do not have distribution rights & every sale is a 1 off & must be approved by the Landlord/Property Owner to gain access to the risers. Just reporting on on-net buildings as done so far in this industry is a hollow fact.

      We must realize that the battleground we are discussing “on-net buildings” is where the ILEC’s defeated the CLEC’s” I call it the Little Big Horn of the 1996 Telecom Act. It was the combination of 2 major mistakes – 1st the ILEC’s received regulatory approval/relief to move the DMARC to the MDF (no longer required to deliver it to where the customer/tenant wanted it). Their argument was that the old inside copper plant on their balance sheet kept them from being competitive with the “CAP companies” fiber plant. This transferred ownership of the ILEC riser infrastructure to the building owner, knowing full well they were 100% on-net & all the new competitive telecom companies would have to get approval from the landlord to gain access to the tenants. Even though these risers qualify as easements these buildings are private property & CLEC’s were unable to gain access under any Federal tariff or litigation. The ILEC’s don’t pay for access but CLEC’s do. The average time to secure a building access agreement is 7 months (BOMA stat). Think about it, the LL views this as a liability & the CLEC views it as an asset i.e. revenue thus placing them at cross purposes. The LL spends more money on building engineering (verifying the riser routes/capacities), co-coordinating with the tenant, & with an attorney than the CLEC agreement will deliver to the LL in revenue in a year or 2. This relationship is a low priority for the LL & a high priority for the CLEC unless the tenant is up for lease renewal. This is a huge disadvantage for the entire CLEC industry.

      The 2nd major devastating mistake made was that all CLEC’s joint trenched on all fiber routes (long haul/metro loops etc) accept for the last mile – guess what that’s the most expensive, longest ROI & timeliest component of the infrastructure build out in all the plant on the books.

  • CroweLoves TOFU says:

    Crowe is master of trying to change the metric. Ask yourself how much profit/MRC you can get for 100 MB circuit in NYC versus the same sqare footage in Peoria. The answer: Peoria is more valuable on a competitive front which means higher MRCs, lowers costs and likely a better price elasticity. Crowe is smelling his own gas again.

    Rob, please don’t go changing, to try to please him. We love you just the way you are. (Billy Joel Crowe).

  • John says:

    Logix Communications……Texas and Oklahoma

    500+ building on-net

    4 million + Sq Ft on-net

    I do agree that you need to look at tenant number as well as number of customers. Also, look at your focused customer. Companies such as TW Telecom and other larger BLEC providers seem to focus on the larger higher bandwidth tenants. Other CLEC/BLECs like Logix focus on the smaller tenants…..More churn but higher profits…

  • Carlk says:

    There goes enron pumping that “little” company comprised of tax cab drivers of his again. One which the revered Oracle’s newest fund mgr. has taken much stock in. You couldn’t make this stuff up if you tried.

    Yes, Jim Crowe would probably piss James Baker off alright, because “it’s not fair to change the rules of the game in the middle of the game,” he once told Supreme Justices and rank and file American listening in order to keep Bush at the head of the class!

    The big question for all of us remains, however. Why is it that the combined (3)/Global Crossing networks will carry more than 50 percent of the globe’s internet related traffic, but can’t produce copious free cash flows nor profits to the satisfaction of Wall Street over their $37.5B investment plant in thirteen years still counting?

    Are we just unlucky? Ask Moe Green!

    It goes beyond some of the usury rates they have been forced to pay over the years during the lowest interest rate period in the history of mankind, and is somehow connected to “square footage” in the “right buildings” positioned in the “right places,” but possibly still the “wrong time” as a result of “regulatory schemes” left over from the dinosaur age when copper ruled!

    Get me that AD revenue attached to my pipes distributing so much crucial DATA X the globe! I for one, have had enough of the BS, and the excuses!

  • Clevus says:

    For what it is worth, you can get some square footage data off of Google Earth for some areas. I see the data in the NYC area and Denver as well. Just access the Earth Pro data set.

  • Anon says:

    Hard to capture, but “buildings” of same footage are all very different… Would rather light a 100,000 foot Colocation data center than a 100,000 foot office building. The 50 largest data centers in USA aren’t one-tenth of 1% of the square footage but account for a substantial portion of carrier revenue (less so for clec’s, to the extent still solvent)

    • Carlk says:

      Anon, you seem to be stating what Jim Crowe is implying in his remarks along with his quest to make this new stat available from (3).

      It dovetails perfectly with Hammerdinger’s two prong explanation of CLEC mistakes including the Battle of Little Big Horn, especially the line of demarcation.

      On the other hand, there is one thing that Hammerdinger puzzled me with tied to his point two being “joint trenching” at least as respects long haul routes.

      I assume that would have been for a limited bunch of large backbone providers in the U.S. originally, and is more confined today because of consolidation to (3), Qwest, AT&T on portions, maybe, and who else, Mr. Hammerdinger? Sprint would have preceded all of that long haul building in the eighties with their now antiquated pin drop LD network. tia

  • Manu5 says:

    Anon,

    Therein lies the conundrum- Lighting a 100,000 square foot data center where you get to sell $2000/mo. 10 GB waves or lighting an office building where you get to sell $2000/mo. 20 Mb circuits with less than half the costs of lighting the data center. Not sure which one I would take!

Leave a Reply to Clevus

You may Log In to post a comment, or fill in the form to post anonymously.





  • Ramblings’ Jobs

    Post a Job - Just $99/30days
  • Event Calendar