Wednesday, August 13, 2008

Telecom Ramblings - Internap, Its CDN, and Strategy

Rob Powell has a nice article about Internap CDN on his blog - basically an answer to an analyst's speculation that Inap could drop CDN in the close future.

It's hard to disagree with Rob, but someone has got to play the devil's part... ;-)

>>"I do think that Internap has to make some decisions from here, but I disagree with the analyst who said that the assets never made sense in Internap’s hands. From a strategic standpoint, a CDN product has always made sense alongside Internap’s high end blended transit and colocation. That’s why they were reselling Akamai’s services before this debacle even started - their customer base really does have interest in CDN services. What Internap has failed at (so far) is execution, not strategy. They have to convince both themselves and their customers that they know how to run a CDN, else they need to sell out to someone who can."

I agree that it made strategic sense for Internap to have CDN - they already had, as a reseller, the best one, AKAM, and they decided to have their own. I may disagree, as an investor, on the acquisition target, Vitalstream , the price paid, etc. - but that's history.

It seemed, after the acquisition, that CDN was the only strategy going forward for Internap - as if growth could only come from that area. Philip N. Kaplan, from Vitalstream, became Chief Strategy Officer for Internap, people like Eric Klinker, who had contributed to the patents that still represent Internap's best assets, left the company at that time.

What I slightly disagree on is that it was only lack of execution. Internap put all its strategic focus on CDN. If you just read their latest annual report, you'll still notice CDN is mentioned twice as much as the other business segments... That to me was choosing the WRONG strategy for the Company - which worries me even more than suffering an integration problem or having inherited a bad customer base.

The real growth area, colocation, was kind of put on hold:

Internap to Invest in Expanded Colocation Facilities

Business Wire, June 12, 2007

Internap Meets Increasing Customer Demand for Premier Data Center Services

ATLANTA -- Internap Network Services Corporation (NASDAQ: INAP), a global provider of optimized, reliable end-to-end Internet business solutions, today announced that it has approved an investment of up to 40 million dollars to fund the expansion of its colocation facilities in several key markets. The company anticipates implementing the expansion over the next three to four calendar quarters, with any potential funding to be provided under standard commercial financing arrangements.

Let's have a look at what happened:

Seattle opened in the 2Q 2008 - 3 full quarters after the announcement.

Houston in the mid of 3Q 2008 - 4 full quarters and a half after the announcement

New York will open in the mid 4Q 2008 - 5 full quarters and a half after the announcement

Boston will open at the end of the 4Q 2008 - I think we can say roughly after 6 full quarters, if on schedule.

We now find out from the 10Q and the conference call that:

>>The increased use of partner sites is one reason for the higher percentage of direct costs for the three-month period ended June 30, 2008 compared to the same period in 2007. We continue to expand the sites operated by us and expect to have more of this space available to be used in the future as part of our data center growth initiative.

We seek to optimize the most profitable mix of available data center space operated by us and our partners.

We expect our recent data center expansion will provide us lower costs per occupied square foot in future periods, enabling us to increase revenues compared to relatively lower direct costs of data center services.

The initial operating costs, especially for rent, of sites operated by us causes us to recognize some costs ahead of revenues but overall is more profitable at minimum levels of utilization than the use of partner sites.

It does sound a lot like: we wish we had had room in our centers, but that wasn't the case... so we had to go for partner sites where our margins are much lower... as also said during the c.c.:

>>George E. Kilguss, III

With regard to your question on margins data center really has two components to the margin. One is the new facilities coming on line. For example Boston, our expansion in New York, those facilities while we’re building them out we’ve taken possession of those and we have to include that rental expense in our cost of goods sold line. At the present time while we’re waiting for some of those factories to come on line we have continued to expand in our partner sites. As I think you know our partner sites’ gross margin is less than our company controlled sites so as we are continuing to sell under the partner sites that also has a natural downward pressure on our gross margins. But the larger component today is this fixed charge that we’re seeing ahead of revenue so once these sites open up we should get a quick improvement because the costs are already in our cost of goods sold and revenue will come in and help offset those costs directly.

(transcripts by Seeking Alpha)

>>"It may make sense for a CDN, one that wants to stay independent of the telecoms, to buy Internap. Hold on, am I insane, you ask? Well think about it for a second. Limelight already resells some IP transit, might it make sense for Limelight to combine its CDN with Internap’s high-end blended transit niche plus a colocation business that would have synergies for all those servers? There are cost savings to be had there. Yes, I do realize that Limelight probably is not in a position to pull it off right now, but I think that is the only big thing in the way of such a move. What about Akamai? If they don’t need a backbone but do need a broader portfolio to compete with one, then Internap’s customer list would not be a bad place to start.

Ah well, I doubt that will happen, but you never know. From here I think all Internap will simply try to regain its reputation, one connection at a time. What do they have to lose?"

Now, I really do not agree.

I just can't see any reason for Limelight to buy Internap at this stage. They don't need the (very limited) revenue generated today by Internap in CDN, and mostly the footprint would be a doubling (Equinix data centers around the world) - I can't see a real advantage.

Limelight would only be distracted by Internap's core business (IP Services...) and I doubt they could benefit from the Internap owned data centers in secondary locations like Atlanta, where the majority of space is (and where Internap is expanding into QTS data center, by the way...) or Houston...

As to Akamai, it seems to me they never believed in the "super bandwidth" proposition, and again I can't see a strategic fit. Not even eliminating a high profile competitor... which Internap, today, is not...

Internap may be an interesting acquisition target, but I doubt because of its CDN business - unless it is an excuse and the rest fits even better in the buyer's business model.

Just as a last comment on strategy, the real issue I have with the Company, as I believe the execution problem in the CDN business are behind us, is the answer to the Italian question "What will you do when you grow up?".

What's Internap, going forward? Still a bundled solution offering, just with a more balanced focus on ALL three products, colo, IP Services and CDN? Opening new locations in Europe looking for new sources of growth in the IP Services segment? Offering data center solution in these markets as well? Partner sites or owned?

Lastly, I was looking at this old chart from an Equinix Investors' Meeting.

Look where Internap stands in the colo business. If we look at Enterprice Value, both Switch and Data and Terremark are now valued about $ 600.000.000, Internap just slightly over $ 110.000.000. Mr Market is often wrong, but I guess this Company made more than just disappoint on execution...

No comments: