The results, so far, have been bit perplexing. After all, we can't publish a list of 1,000 teen-targeted mobile content companies and 100 user-generated content Websites.
Indeed, there will be some application and content companies on our list. What about infrastructure startups? The first problem, of course, comes with the word “new.” Many of the “startups” we found, were in fact, at least three to four years old. And there were very few infrastructure startups formed in the last 24 months.
Mostly we had to look in places other than core telecom technology to find the most recent startups. The thing about those clever venture capitalists, of course, is they always seem to have the same idea at once. Once you expand to the realm of digital media or wireless applications, the list of startups explodes.
Why venture capitalists always move in such packs is a mystery to me. Maybe it’s that VCs really are the same as Joe-Q public investors – they like to migrate where the latest fad or momentum is. Herein may lie the secret to why certain telecom infrastructure startups, rather than the faddish user-generated content startups, might be the ones to watch here.
After all, how much innovation comes from so-called "user-generated" sites, such as YouTube. YouTube is useful and entertaining, but I have a hard time seeing it as the next Disney Corp. Primarily, the site specializes in distributing thousands of hours of jittery video clips, much of it belonging to someone else. Imagine your cable company filling up your channel lineup with hundreds of community access channels. That's YouTube.
What about real enabling technology? The kind that, rather than filling up pipes with grainy footage of fraternity videos, can reduce distribution or delivery costs by a factor for 10 or more? There doesn't seem to be a whole lot of investment in this type of stuff going on.
I’ve mentioned before that the telecom investment cycle tends to run in about 10-year patterns. This is how long it takes service providers to exhaust the older technology and require something new be put in place.
The last major run on networking infrastructure technology, as we know, was in 1999-2000, when carriers underwent massive upgrades of their fiber optic backbone infrastructures. You could argue that another investment cycle got underway in 2004/2005 in the access market, as the carriers focused on hooking up the pipes to the customers that would feed this massive core infrastructure upgrade.
So what should we look for in mid-2006? It now turns out that very few infrastructure startups of any kind are being started up. Think of the last time a cutting-edge core router startup was founded. Can you think of any? There was Procket, but that company was incorporated in 1999. (See Procket Reaches 'End of Life'.) What about edge routing, optical switching, or even wireless backhaul? I’d guess that the number high-profile startups in these markets in the last two years totals less than ten.
Where does that leave the carriers in 2008-2009, when they will arguably be ready for new technology that (1) helps alleviate new bottlenecks in the core from the dozens of user-content creation and wireless applications companies the venture capitalists have funded; and (2) helps the carriers get ready for another core infrastructure upgrade cycle?
The answer, of course is: Deep doo-doo. There won’t be any new core infrastructure technology!
A prime example of this phenomenon is Infinera Corp. (Nasdaq: INFN), one of the last great infrastructure startups. Infinera has figured out how to lower the economic barrier to upgrading to 100-Gbit/s optical links. Apparently very few companies have been working on this.
Nearly every carrier technologist I’ve spoken to in the last three months has mentioned Infinera, and how they love the technology. They don’t mention anybody else. Why is that? Well, first of all, apparently Infinera has some very attractive technology. But more importantly, it appears to be the only company that’s been working on a big, big infrastructure problem! Infinera is the lone wolf of core optical newtorking – and it's going to enjoy the spoils in what has been, remarkably, an uncrowded space.
The thing is, Infinera, which is prepping for an IPO, isn’t really even “new” anymore. It’s moved well beyond the startup phase with millions of dollars in sales. Amazingly, Infinera was started back in 2001! That was a sort of dead zone for telecom investment as the bubble came crashing down. But what’s next after Infinera? There appears to be even less new technology on the horizon. In fact, I can't think of one interesting optical networking company formed in the last three years. Can you?
Now, imagine where we are four or five years from now. The world will be deluged by 2,000 moblie content applications and 1,000 user-generated content Websites, including 50 knock-offs of Youtube.com. The only problem? The infrastructure won’t be able to handle it, because no new enabling technology startup techology is being developed, and the R&D budgets of large companies have been slashed to the bone.
Get to work, startup world. Meanwhile, if we're missing something in that world, please drop us a line at [email protected].
Our list of "Top Ten New Startups to Watch" will be published this week.
— R. Scott Raynovich, Editor in Chief, Light Reading