Main | July 2004 »

June 24, 2004

AdTech notes from John Zagula

If you care about internet marketing, read this post. John Zagula posted this earlier, but has agreed to have it cross posted here.

AdTech Forum - Is the bubble back?

Just got back from AdTech Forum. Was supposedly much bigger, hypier and more well funded, and attended than before.

But hey, at least I brought home some nice tchotchkes. My kids especially liked the flashing bouncing ball from Who’s Calliing, the weird belt clip light from Atlas Search and the frisbee thing from Digital Envoy. The little flashing light thingee from Google was cool but too sharp for play. And honestly the little sticky-note book from AOL was not good enough, they should do better. Oh, yeah, and I didn’t win ANY of the many raffles for free iPods and TiVos. What a rip. (Strange, I got much better swag from an insurance industry conference I went to recently)

Oh, but I guess that was not what I was there for. Sorry. I’ll get to the actual content now.

Who was there? Tons of vendors like DoubleClick, Revenue Science, GoToast, CasaleMedia, Atlas DMT. Consultants and agencies like SBI Razorfish and Media owners, aggregators, brokers etc like Claria, Tribal Fusion, Fastclick, and Kanoodle, and of course, the company everyone wanted to be like, Google. Just to name a very few.

The Big themes: accountability, integration, the death of TV, contextual vs. behaviorial targeting, pay for performance, measurement, measurement, measurement - CPA vs. CPM vs. ROI vs. eieio, other stuff like promotional and other marketing vehicles. And most importantly, why is online still such a small % of the total ad spend?

What’s Hot supposedly: analytics, consulting $, SMS marketing, DVRs shaking up the world, metrics, mixed media schedules and planning, broad-reach/interactive integration and planning, local, email, search and more search

What’s Not supposedly: interactive TV, streaming ads, promotions.

See the VC discussion for more on this. Everyone of course wanted to know what Mark Kvamme of Sequoia (Yahoo and Google of course) had to say. Which was really good, but the other panelists had a bunch of other observations. Like deal prices are starting to get scarey again, also that we've really only seen the beginning of the shift to online ad spend, about how important local is but how high the cost/advertiser acquired. About how wireless is super cool and could leapfrog PC/internet for streaming video.

Biggest Gap: strangest thing to me at least was how little attention blogging as a business opportunity got, even after all the Billg talk. Most pooh poohed it as not investable, needing an ad model, and needing editorial quality control. Humph. Well, at least Mary Hodder found some enthusiasm about blogs there.

Other, final impressions: Lots of buzz words like “new information economy” “hyper-changing” “leverage.” Lot's more optomism and excitement but I sensed an undertone of Vienna before the Anschluss, waiting to see if the shoe drops. Really nice to see a guy like Mike Windsor take the whammy out of a bunch of the claptrapier stuff said. Favorite quotes are from him.
"It's about money. Remember that. So then we have to go where the money is. The billion dollar not million dollar budgets. And they are with broad reach clients not interactive clients. Until interactive is a significant and regular part of the ad program we will always be fighting an uphill battle." Also "Online represents less than 10% of the budget but more than 90% of the analysis. Why, because it can be analysed. Not because the analysis really tells you any more about the true ROI of the effort or how it performs relative to anything else."

Posted by Martin at 2:57 PM | Comments (0) | TrackBack

June 23, 2004

The VOIP Man Commeth

Many of us in Northwest we have personally felt the telco meltdown since 2000. My friend Greg Maffei at 360 Networks may be one of the few to navigate the storm albiet at the cost of taking his company through a wrenching bankruptcy. ATT Wireless (in Redmond Wa) just got swallowed up. TMobile is tossing around for a model that works (I still will never pay for their hotspots at Starbucks). The other shoe is about to drop on these guys though and it is VOIP.

Russell Daggatt, local entrepreneur and veteran of many McCaw companies has recently written the most cogent analysis of the recent past and near future of telcom. The bottom line is that you and I will benefit if the government doesn't screw it up (paraprasing here). He published this as a special letter in Mark Anderson's Strategic News Service. Russ has gratefully agreed to have the whole thing re-printed here. Thanks Russ for your contribution!

This May 12th, 2004 Issue:

***SNS*** Special Letter: Internet Telephony: Storming the Bastille

By Russell Daggatt

Provided by: Technology Alliance Partners
On the Web:


By Russell Daggatt

Act I: So Much for the '96 Telecom Act

Mark has taken to calling them the "Nasty Babies." And not without cause. These offensive offspring of the once monolithic Bell system monopoly are known in more sober industry parlance as the Regional Bell Operating Companies (RBOCs). If you live in the U.S., there is about a 97% chance that you get your local phone service from one of them. In other words, they are monopolies. And they act like it.

The 1996 Telecom Act was supposed to create competition by offering the RBOCs new worlds to conquer in return for opening up their local fiefdoms. But the RBOCs stonewalled, through litigation and brazen flouting of the law, until the nascent competition had all been sent to debtors' prison. (Since 1996, the RBOCs have been assessed over $2 billion in fines and other similar regulatory levies - a small price to pay for preserving their monopolies, don't you think?)

Remember the Competitive Local Exchange Carriers (CLECs)? These were the new companies spawned by the '96 Act which were going to compete with the RBOCs. Allegiance, Covad, Focus, GST, McLeod, NorthPoint, Rhythms NetConnections, Winstar, XO. These, and many more, all went bankrupt. At their peak in 2000, the CLECs as a group had a market cap north of $85 billion. By the end of 2002, that had shrunk to just over $1 billion.

The Interexchange Carriers (IXCs) haven't fared much better. Worldcom, Global Crossings, Williams, 360Networks, and others. All belly up. Before the meltdown was over more than 100 telecom companies had gone bankrupt.

Even AT&T, once the Mother of All Monopolies, is limping along on stubs. Like an animal caught in a trap, it has been gnawing off its limbs - first cable and now wireless - just to survive. (PanAm did the same thing in the airline industry back in the '80s, selling off international route rights piecemeal to generate operating cash. Remember PanAm?)

If the CLECs and IXCs weren't able to challenge the RBOCs, then at least the RBOCs would cross over into each other's markets -- Or so the theory went. Instead, like a game of Pacman running on the Kanagawa Earth Simulator, the RBOCs swallowed up each other. GTE, Bell Atlantic, and NYNEX eventually became Verizon. Southwestern Bell, PacTel, Southern New England and Ameritech became SBC Communications. Before the '96 Act, the four largest local exchange carriers owned fewer than 48% of the lines in the country. Now that number is over 85%. Verizon is also the country's largest wireless carrier, soon to be overtaken by Cingular (owned by SBC and BellSouth) when Cingular digests AT&T Wireless.

In the immediate aftermath of the '96 Act's passage, many hopeful young bandwidth lovers thought that the competition to the RBOCs might come from cable providers. AT&T's CEO, Mike Armstrong, got caught up in this romantic notion and, in 1998, shook the telecom world by announcing AT&T's betrothal to cable giant TCI in a deal then valued at $48 billion. AT&T was going to upgrade the TCI cable system to get into the local phone business. But when the pheromones wore off, Armstrong woke to discover that TCI's John Malone had cleaned out his bank account and left him a bunch of junk. (Malone kept the good stuff, TCI's programming assets, which were spun off as Liberty Media.)

AT&T also bought cable provider MediaOne for $62 billion. Unfortunately, the economics of providing switched telephone service over cable TV systems never worked out, and in 2001 AT&T was forced to sell off its cable assets to Comcast for $52 billion - about half what it had invested in those assets.

Meanwhile, the pattern that played out with local phone service has been pretty much the same in the cable market. The major cable companies have not invaded each other's service areas. Instead, they have merged into huge dominant national firms with tightly-controlled clusters of properties. There has been almost no "overbuilding" of new systems. And the major incumbent companies control about 97% of the wireline multichannel video market, virtually the same as the RBOCs' share of the local landline voice market. (While satellite TV offers credible competition for multichannel video, with about a 20% market share, increasingly the cable to the home is taking on a significance that goes beyond just the video market.) Fewer than 5% of American households have a choice of cable providers. The result has been cable TV rates increasing since the '96 Act at three times the rate of inflation.

Oh, well. So much for the '96 Act increasing competition.

Act II: VoIP to the rescue

But, wait. There may be hope.

The local voice market has been the RBOCs' fortress. Local voice and switched access revenues generate at least 75% of their profits. But it is no longer enough for them just to defend that fortress. After years of access line growth as people added faxes and computer connections, the number of lines now is actually decreasing. Mobile phones are replacing second or third lines - in some cases, even the primary line. (Hard figures are elusive, but one recent survey showed that 3% of Americans have dropped their primary landlines and 8% expect to follow within the next five years.)

But the real competition is likely to come from Internet telephony, generally referred to as "voice over Internet protocol" (VoIP) or "voice over net" (VON). Switched service may not be economic over cable, but VoIP through a cable modem is another matter. It costs a cable operator as much as 75% less for an Internet phone hook-up as it does for a circuit-switched connection.

Disruptive technologies tend to be overestimated in the short term and underestimated in the long term. And make no mistake about it: VoIP is a disruptive technology. It changes the fundamental business model for the biggest segment in the telecom market: fixed-line voice service.

Prophesies of VoIP as the future of telecom have been widespread ever since Netscape's browser launched the Internet into the mainstream a decade ago. But early versions of VoIP technology were pretty primitive, using speakers and a microphone attached to a PC. Its appeal was hindered in those days by the Internet's limited bandwidth, high latency and overall low quality of service. Many early adopters were foreign students and immigrants using VoIP for very cheap (and poor quality) international calling. Even today, VoIP usage is greatest for international calls where by some estimates it constitutes over 10% of the traffic. [Pub. Note: VoIP has also achieved good penetration in the relatively "invisible" prepaid cellular and LD markets, as noted here earlier.]

VoIP is still just an insurgent threat at the moment. The largest providers have subscriber counts only in the small six digits (out of 182 million local access lines in the United States). But it is gaining momentum. There are countless small startups focusing on one aspect or another of the VoIP market. VoIP provided in fairly conventional packages through service providers and connecting with the public switched telephone network is known as carrier VoIP (to distinguish it from peer-to-peer VoIP). Among other things, carrier VoIP allows you to use a regular phone and a conventional phone number (pursuant to what's known as the North American Numbering Plan, or NANP).

Among the larger new carrier VoIP service providers is Vonage, which claims to have more than 100,000 residential and small-business VoIP customers in the U.S.. AT&T is rolling out VoIP service under the brand name CallVantage, and is claiming it will be in 100 markets this year.

Cable providers have been slow to roll out their own VoIP services, but they are starting. Time Warner launched phone service in Portland, Maine, last year and has already signed up over 150,000 customers (more than 10% of its cable subscribers in that market). It plans to roll out the service to all of its 10 million customers by next year. Other big cable providers are similarly getting into the game.

Mobile carriers began the trend toward pricing models that are independent of distance. Now VoIP adds to that a decisive break away from pricing based on minutes of use. Most carrier VoIP providers offer unlimited local and long distance calling for a fixed monthly price, generally around $40 / month. (International rates are still billed by the minute. They vary among providers but are generally $0.02 to $0.10 / minute for major Europeans countries and between $0.10 and $0.30 / minute for most other countries.)

Another attraction of carrier VoIP is the almost unlimited feature set. A residential user or small business can have access to features that rival or exceed those of the most sophisticated PBX or Centrex systems. Voice mail (which you can forward via email), caller ID, call waiting, conferencing, call blocking, and call logs are among the features that are typically bundled into the fixed price. You can have your own "hold" music or have an auto-attendant answer your phone. One phone number can ring in sequence or simultaneously to multiple phones.

In the enterprise environment, a large number of users can share a limited set of phones with each user taking his own phone number with him to whatever phone he is using across multiple locations. Best of all, multiple users at a single site can all share one data line rather than having to pay for individual voice circuits. (If you want to see a big decrease in access lines, just wait until VoIP really catches on in the enterprise market. That would be soon.)

For this mainstream variety of VoIP you need a broadband connection and a digital adapter through which you can attach any standard wired or cordless phone. (Vonage says it will offer a WiFi phone later this year, which you can use to connect at home or through any WiFi hot spot elsewhere. Dual-mode cellular / WiFi phones are also coming onto the market.)

Number portability is still something of an issue. In theory, you can transfer your mobile number to a VoIP account. But if you are using DSL for your broadband connection you need to keep whatever number you are using for that service and get a new number for your VoIP account. If you do try to transfer your current landline number, you can probably assume the RBOCs will make it as painful as possible.

In brief, that is carrier VoIP. If circuit-switched telephony is yesterday, carrier VoIP is today.

And Skype is tomorrow.

Act III: Think Skype

Started by the same Scandinavian duo, Niklas Zennstrom and Janus Friis, who created KaZaA, Skype is based on the same peer-to-peer (p2p) technology as that hugely successful file-sharing system. Like KaZaA, it consists entirely of free software with no central administration or infrastructure. It avoids the problems of scaling by joining together all nodes in the network to dynamically participate in traffic routing, processing and any bandwidth-intensive tasks that would otherwise be handled by central servers. New nodes in the network provide the processing power required to scale the network in proportion to those new nodes.

Skype appears to have overcome the sound quality and call completion problems that plagued early p2p VoIP efforts. Those pioneering technologies also had problems getting calls past firewalls and the use of network address translation (NAT), which rendered most residential computers unable to communicate with traditional VoIP software. Skype has solved those problems without the use of centralized software or directories. It uses "supernodes" that communicate in such a way that every node in the network has full knowledge of all available users and resources with minimal latency. When a user logs on, all other users immediately have access to that knowledge. Skype intelligently routes fully-encrypted calls along the most effective path. It keeps multiple connection paths open in order to dynamically choose the one that is best at any given moment.

At present, Skype still operates only on computer-to-computer connections, but the company is working on connections to conventional landline and mobile phones. In theory, you don't need to use a broadband connection - Skype only uses between 3 and 16 kbps - but you are more likely to encounter service problems if you don't. With a broadband connection, however, the service is little more than noise in the bit stream. Too cheap to meter, as the nuclear power promoters used to say.

In other words, it is for all practical purposes free voice service. Since the Skype software was introduced last August, there have been more than 10 million downloads - still a drop in the bucket in a world of a billion landlines and another billion mobile phones. But it's not hard to see where this is going.

Zennstrom has said that when he and Friis started Skype they were looking for a p2p application that would have a "major disruptive impact." How about eliminating the premium for voice service (what Zennstrom calls "rip-off pricing")? I think that qualifies. The RBOCs' fortress is going to crumble like a sandcastle at high tide.

None of this is going to happen overnight. But when imagining the future of telecom, and contemplating the myriad economic and regulatory issues surrounding it, think of everyone having a big dumb broadband pipe for a flat fee. Think of voice as just another (fairly-inconsequential) bit stream. Think of most applications relying on some variation of p2p architecture to take advantage of cheap, distributed, easily-scalable processing power.

Think Skype.

Act IV: Waiting for Broadband

The key to all of this is broadband penetration. As Mark points out often and emphatically, the U.S. is a laggard among major industrial nations in that regard, ranking only 11th worldwide in broadband penetration. And that is virtually all cable modem and DSL service.

By contrast, in February of this year Japan - a country with less than half the population of the U.S. - quietly marked a major milestone as the millionth subscriber was connected with fiber-to-the-home service. (It was only in March of 2001 that fiber-to-the-home was first introduced in Japan by USEN Corp.) I am not aware of anyone, other than Bill Gates and a few residents of Palo Alto, having fiber-to-the-home in the U.S.

In Japan, VoIP has also played a big role in the deployment of broadband. For example, Yahoo Japan rolled out its broadband phone service just over two years ago, in April 2002, and already has 4 million users. The biggest U.S. providers still have fewer than 200,000 subscribers.

The story is even more dramatic in South Korea, where two-thirds of households have high-speed connections and those connections are generally from 4 to 64 times faster than the average "high-speed" connection in the U.S.. (South Korea, with a population of 50 million, also has 30 million mobile phones, of which 25 million are Web-enabled.)

For those keeping score, Hong Kong, Canada, Taiwan, Denmark, Belgium, Iceland, Sweden and the Netherlands also top the U.S. in broadband penetration.

But, you might be wondering, what happened to all of that telecom investment made during the glorious bubble years? The companies may have gone bankrupt, but the assets they created are still out there somewhere, aren't they?

Indeed, it's been estimated that between 1997 and 2001 over 100 million miles of optic fiber were laid worldwide. That's enough fiber to reach the sun. Unfortunately, it doesn't reach the homes of most consumers. At least not in the U.S.. There are now a lot of sub-sea and long-haul fiber cables, and multiple overbuilds of the fiber rings surrounding major metropolitan areas, all averaging around 52 fibers per cable. About a third of that build-out was in North America. But by the time the bubble burst, less than 3% of the fiber was actually in use. (At least, you don't have to worry about long distance rates going up any time soon.) Meanwhile, you and I still have to rely on our coaxial TV cable, or worse, our local RBOC's copper phone wire.

As of last month, 48 million American adults, or 24% of the adult population, had broadband Internet access at home. That is a 60% increase since March of 2003. Until recently, the home broadband story has been almost entirely about cable. A year ago, two-thirds of home broadband connections were cable modems. What has been most significant since then has been the surge in DSL connections. They more than doubled and have been largely responsible for the big increase in the growth of home broadband access during the past year. DSL now has a 42% share of the home broadband market, up from 28% in March of 2003. (Wireless and satellite have been holding steady at around 4%.)

It seems the RBOCs are finally starting to get the picture. As long as they could survive and prosper by defending their "rip-off pricing" of voice service, they had little incentive to roll out DSL. To the contrary, it threatened to commoditize bits and ultimately undermine voice revenue. But the combination of mobile phones and cable modems is now doing that anyway. To replace the loss of voice revenue, the RBOCs have had to accelerate their DSL build-out. As VoIP begins to really take a bite out of voice revenues, that imperative is only going to increase. But as streaming video starts to catch on, U.S.-style DSL just isn't going to cut it. At some point, if the RBOCs really want to stay competitive with cable (in other words, if they want to survive), they are going to have to start building out fiber to the home.

While the long-term dominance of VoIP is all but inevitable, it would be nice if it happened in our lifetimes. The faster the voice fortress is eroded, the sooner we may actually see fiber in our future. The biggest threat to that future is regulation. VoIP is an irresistible force, like gravity. But if VoIP is gravity, legacy regulations are inertia.

Act V: The U.S. Regulatory Labyrinth

Conventional phone service is subject to all manner of federal, state and local regulation, taxes and fees. There are regulated tariffs and access fees. There is the Universal Service Fund (USF). There are requirements for 911 emergency services and wiretapping capability. While much of this burdens the RBOCs, they are also the beneficiaries of required access fees from the IXCs and wireless carriers.

By contrast, the Internet has prospered in a regulatory-free zone. From the beginning, ISPs were exempt from virtually all regulation. And the RBOCs (unlike the cable companies) couldn't prohibit or burden ISP access to their networks. In that environment, Internet access flourished in this country. Thousands of ISPs promoted narrowband Internet, selling it at prices that were one tenth of what was charged in Europe and one hundredth of what was charged in Asia. This policy of regulatory forbearance is a big part of the reason why the Net really began as an American medium.

But if the RBOCs know anything, it is how to game the system. They would love to subject carrier VoIP to the same labyrinthine regulatory environment that applies to conventional phone service. And not just RBOCs, but virtually all the incumbent carriers are threatened by this disruptive new technology. If, in the words of Samuel Johnson, patriotism is the last refuge of a scoundrel, regulation may be next to last.

So far, the incumbents have been largely unsuccessful in their regulatory efforts to block or hinder VoIP. As far back as 1996, a trade association of 230 telecommunications service providers unsuccessfully petitioned the FCC to ban VoIP. And as recently as February of this year, the FCC ruled that the Free World Dialup (FWD) service is not subject to regulation by the Commission. (FWD is generally considered the first IP telephony network. It was started as a voluntary computer-to computer network consisting of servers throughout the world to provide free or very cheap long distance service to anyone who wanted to join the network. Think of it as sort of an early spiritual ancestor of Skype.)

In a series of decisions over more than 30 years, the FCC has drawn a distinction between "basic services" and "enhanced services." That line of decisions established that the enhanced service market was highly competitive with low barriers to entry; therefore, providers of those services would not be deemed "common carriers" subject to myriad FCC regulations. Among the regulations to which "enhanced services" are not subject are those dealing with tariffing, service nondiscrimination, interconnection, disability access, privacy, wiretapping access and contribution to the USF. (The FCC continued to impose conditions on entry into the enhanced services market by established telecom service providers and on their provision of basic service to enhanced service providers.)

That distinction between basic services and enhanced services was mirrored in the '96 Act. But to keep it all from being too easy for us to understand, the '96 Act uses different terms: "telecommunications services" and "information services." The FCC has interpreted the '96 Act's distinction between those two categories of service and the Commission's own distinction between basic and enhanced services as being essentially the same thing.

For all the failures of the '96 Act, it had one great virtue, summarized in the following policy statement: "It is the policy of the United States ... to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal and State regulation. . . ." Reflecting that policy, it is now well established by FCC decisions that ISPs are providers of enhanced / information services and are free from access fees and from obligations to pay into the USF. It is also clear that some VoIP falls into that regulatory-free zone, as well.

On the other hand, the FCC has not gone so far as to say that any IP in the network, no matter how insignificant, makes a call VoIP, free from regulation. Less than three weeks ago, in a fairly limited decision, the FCC ruled that AT&T could not avoid paying access fees by virtue of routing some portion of a regular circuit-switched call over its IP backbone.

From all of this it is clear that computer-to-computer VoIP (whether carrier or p2p) is clearly an enhanced / information service free of regulation. It is also reasonably clear that a service that starts and ends in a circuit-switched connection is a basic / telecommunications service, even if it happens to be translated into IP packets somewhere along the way. Things start to get murky when a call involves a circuit connection on one end but not the other.

How much IP makes a call VoIP? Does the use of a traditional handset, with a NANP phone number, subject a call to regulation even if there are no circuits involved? The FCC currently seems to be treating as the decisive factor whether there is "net protocol conversion to the end user," whatever the heck that means. My guess is that, in the near term, the distinction will turn on whether the call starts or ends with a circuit connection. But even that distinction doesn't really make much sense long term.

The problem with attempts to apply the legacy regulatory structure to VoIP is that VoIP exists ultimately only in the service layer. Think Skype. It doesn't adhere to any local, state or national boundaries. It's just a distributed software application. By contrast, the transport layer - the physical wires, cables and transmitters - has a physical presence that can be subjected to whatever new regulatory scheme ultimately emerges from the current chaos. You could even just port over to the transport layer most of the regulatory apparatus that is currently being applied to basic voice service.

And in fairness to the RBOCs, as they and the cable providers increasingly come to resemble each other, the rationale for regulatory distinctions between the two means of transport becomes even more incoherent than it already is.

One important example of disparate treatment of phone companies and cable providers is the issue of ISP access to their networks. Phone companies are required to provide nondiscriminatory access to their networks; cable companies aren't. This has allowed the cable companies to roll out their cable modem services using the same closed proprietary model that they use for video services. If you want my cable modem service, you have to use my ISP.

Of course, the phone companies would like to adopt the same proprietary model as the cable companies. But for the sake of consumers, parity should work the other way. Both types of transport should be equally open to nondiscriminatory access by ISPs. You should be able to choose your ISP just as you do your long-distance provider today. This is one of the most important regulatory issues affecting the continued development of the Internet in this country.

And nondiscriminatory access should include no interference with the quality of any particular service provided over the network. There is some speculation in tech circles that the cable companies may try to establish, and the phone companies maintain, a premium for branded voice service over their networks by, for example, introducing a small delay into their networks for all packets other than their own voice packets. A half second delay wouldn't be noticeable for most Internet services but would be unacceptable for voice service.

Consumers should be allowed to run all applications of their choice and to attach any device they choose, as long as they do not harm the provider's network, enable theft of service or exceed the bandwidth limitations of their service plan.

At least a couple of key rulemaking proceedings are before the FCC at the moment. In March of this year, the FCC initiated a proposed rulemaking concerning "IP-Enabled Services," which will probably be the Commission's vehicle for sorting out most of the issues involving VoIP within the existing regulatory structure, such as whether VoIP providers must pay access fees and support the USF. Some of those same issues may also be addressed in another proposed rulemaking concerning "Intercarrier Compensation," which was initiated in 2001.

Ultimately, more fundamental changes in the telecom regulatory structure will probably be required. Toward that end, Congress is beginning to revisit the whole matter of the '96 Telecom Act. While some changes clearly are in order, that doesn't ensure that the changes will be good ones. The scene is certainly ripe for mischief.

In particular, the Republican leadership in Congress has been inclined in the past to give the RBOCs pretty much whatever they want. Fortunately, there seems to be a pretty widely shared view in Congress and at the FCC that the Internet and related information services should be left in the regulatory-free zone. Even RBOC supporters tend to view that precedent as the nose under the tent that could eventually free them from much of the current legacy regulation.

As a general principle, the relevant regulatory distinction should be between the transport and the service layers. The service layer should be entirely unregulated. Regulations and things like USF contributions should be pushed as far as possible "down the stack" toward the transport layer.

And in the meantime, do no harm. The best thing the FCC could do is to clarify that VoIP, at least to the extent that it is untainted by circuits, remains entirely free from federal and state regulation.

Act VI: Storm Clouds on the Horizon

But a new threat to VoIP has emerged that could have broader implications for the infotech world as a whole. It is coming from the FBI and Drug Enforcement Agency. Concerned that the growth of VoIP could undermine existing wiretap capabilities, in March, the Justice Department filed a 75-page petition with the FCC seeking to require that VoIP providers, and other providers of "information services," build into their networks and applications a surveillance capability. That petition would potentially greatly expand coverage of the Communications Assistance for Law Enforcement Act (CALEA).

Enacted in 1994, CALEA required telecommunications companies to redesign their networks to provide access to law enforcement agencies for wiretaps and other surveillance measures. But CALEA specifically exempts "persons or entities insofar as they are engaged in providing information services." The term "information services" was broadly defined to cover current and future advanced software and software-based electronic messaging services, including email, text, voice and video services. In general, CALEA only applies to traditional telecommunications services regulated under common carrier rules. (The one exception to that to date has been DSL. The FCC has ruled that because DSL is provided over transmission facilities that also provide telecommunications services, the transport layer of DSL is also covered by CALEA.)

In essence, the Justice Department's petition asks the FCC to rule that CALEA will automatically apply to any new technology that directly competes with any service that is covered by CALEA. Anyone who introduces a new technology that competes (or might compete) with an existing technology covered by CALEA would be required to go to the FCC to find out whether the new technology is covered. If it is, the technology could not be deployed until a fully satisfactory intercept capability was incorporated into the service. In other words, there would be a "pre-clearance" requirement for new information technologies to the extent that they might compete with the telecommunications services currently covered by CALEA. This could extend to the enabling equipment as well as services.

While CALEA does not currently require that those who provide information services build in specific surveillance capabilities, those service providers still have to comply with a legitimate law enforcement request for records and other information, including surveillance to the extent possible. Law enforcement agencies make tens of thousands of such requests each year, and providers of information services and equipment have committed substantial resources working with those agencies to develop new technical capabilities and procedures to facilitate those requests. Law enforcement agencies can pull from the data streams presented to them by information service providers whatever information they have been authorized to obtain.

The issue is whether information service providers have to build into their networks or applications specific surveillance capabilities and whether the FBI should have pre-clearance authority for those new technologies.

Although Congress permitted the extension of CALEA beyond common carriers, the extension is limited to those situations where the FCC determines that a particular entity is providing a service that is a replacement for a substantial portion of the local telephone exchange service, and that it is in the public interest to deem the service provider a telecommunications carrier for CALEA purposes. The Justice Department argues that a telecommunications service includes not only circuit switching but also packet switching. But such an expansive interpretation would gut CALEA's information services exemption and sweep in virtually all of the Internet.

The effect would likely be that no new Internet services could be introduced without substantial regulatory proceedings and a design that meets all the desired law enforcement requirements, starting with version 1.0. Even existing broadband networks could have to be retooled to comply with CALEA. A regulatory pre-clearance process would delay and potentially prohibit the deployment of new technologies. It also would impose the cost of building wiretap capabilities into new technologies even before it is clear whether they will succeed in the market.

Certainly in the wake of 9-11 people are more inclined to defer to the needs of law enforcement and national security than they might have been in the past. But if you really want to kill information technology innovation in this country, let the FBI dictate how the Internet should be engineered to permit whatever level of surveillance it deems necessary. Remember, this is an agency whose own information system consists largely of file cabinets filled with paper documents. The United States does not have a monopoly on the development of innovative information technologies. Internet traffic travels anywhere in the world and facilities that act on that information can be (and are) located anywhere.

It's hard to know how this will get sorted out. You might recall the fight a decade ago over the Clinton administration's "Clipper Chip." The NSA and FBI wanted to require that all exports of private encryption technology employ a government-designed encryption chip providing U.S. law enforcement and intelligence agencies a "backdoor" to de-scramble communications. After a rough battle, the tech community eventually succeeded in killing that proposal. (The Congressional leader in that battle was Senator Maria Cantwell, then in the House, who will be speaking at the opening night dinner at the FiRe conference later this month.) The issues are very much the same this time around. The national security climate has changed, but the importance of the Internet and other information technologies is also appreciated to a much greater extent than it was in 1994.

My guess is that the FCC will resolve the CALEA issues by pretty much adhering to the existing service category distinction between basic / telecommunications and enhanced / information services. As with VoIP, the key distinction, at least as a practical matter, will probably be whether the communication entails the use of a circuit-switched network.

In any event, regulation, including an expansion of CALEA, threatens primarily carrier VoIP (with or without connection into the circuit-switched networks). If burdensome regulation is imposed on those services, it will only hasten the inevitable movement to p2p VoIP services that don't entail any carrier assistance.

And how in the world do you regulate that? It's a bit like old King Canute placing his throne on the beach and commanding the tide to leave him dry.

Think Skype.

Copyright 2004 by Russell Daggatt.


About Russell Daggatt

Russell Daggatt, 48, is a private venture investor and corporate advisor. He spent over seven years working with cellular-pioneer Craig McCaw in various roles in Seattle and London, including President and Vice Chairman of Teledesic LLC; CEO of New ICO Global Communications, Ltd.; and CEO and Vice Chairman of ICO-Teledesic Global Ltd.

Daggatt is co-author (with Tren Griffin) of the book The Global Negotiator - Building Strong Business Relationships Anywhere in the World (Harper Collins, 1990). He is a graduate of the Harvard Law School and University of Oregon.

Daggatt was selected as a "Global Leader of Tomorrow" by the World Economic Forum in Davos, Switzerland. He has been a speaker at Esther Dyson's PC Forum, Bob Metcalfe's Vortex, George Gilder's Telecosm, and Mark Anderson's FiRe, among other industry conferences. He has been the subject of cover features in Red Herring and Chief Executive magazines, and has appeared on CNN, CNNfn, CNBC, and MSNBC.


I would like to thank Russ for taking quite a bit of time to put together this painstaking analysis. One hopes that our Congressional leaders will consider the ideas laid out here in forming the next steps in policy, so that the U.S. can again accelerate its economic development, instead of slowing down national priorities in order to feed the monopolies. As noted, readers will have a chance to hear and meet both Russ (a FiRe Advisory Board member) and Senator Maria Cantwell at FiRe 2004.

Your comments are always welcome.


Mark R. Anderson

Strategic News Service LLC Tel. 360-378-3431
P.O. Box 1969 Fax. 360-378-7041
Friday Harbor, WA 98250 USA Email:

Posted by Martin at 8:02 PM | Comments (0) | TrackBack

Entrepreneur Success Factors

Just posted a couple of thoughts on this over at: VentureBlog: Thinking About Success Factors

Posted by Martin at 8:21 AM | Comments (0) | TrackBack

June 21, 2004

Tangible evidence in a recovery in Venture

Mondays are very special days for VCs. Usually it is the only day all week that you can catch most of the firm together since it is traditionally the Partner Meeting day. In the meeting we split time between portfolio company updates, office administrative, EIR updates, new company presentations and new investment discussion/votes. Meetings can be 15 minutes to all day events. It is a fairly accurate shorthand for how busy a Venture Capital company is by how long the meetings are and the mix between internal issues and new investment discussions.

Two years ago, I would guess our meetings were running 1-2 hours with about 90% of the time spent on the current portfolio. We had 2-3 new company presentations a month. This time last year, the meetings were about the same length, but the amount of time spent on new investments had risen slightly. Maybe 70/30 portfolio/new investment time and probably 3-5 new companies a month. Six months ago the meetings started to lengthen to 2-3 hours and the mix was more 50/50 with 4-6 new companies a month.

Today, we had a 6 hour partner meeting that was 95% new investment discussions with 5 new companies presenting in one day! Now, granted we skipped one meeting last week due to lots of people out of the office, but the level of activity is still extradornary when compared to the last two years.

Overall I would say that the number and quality of deals we are seeing has picked up significantly over the last six months, much like it has in the Valley. There are studies that track this kind of thing on a regional basis and when the update comes out, I will post it. Bottom line: Now is a good time to start a company and the VCs are actively listening!

Posted by Martin at 10:16 PM | Comments (0) | TrackBack

June 15, 2004

Hello World

Welcome to NWVentureVoice. Our goal with this blog is to discuss the things we think about on a daily basis as venture investors in the NorthWest (yea that place with all the rain and trees). Although Gartner will tell you the the upper left hand corner of the quadrant is never where you want to be, we believe they are wrong when it comes to starting companies. We believe the NorthWest is an exceptional place to start and grow companies. Here we'll discuss why.

As this site develops, we'll add info about ourselves. All the contributors are Venture Capitalist with an interest in the NorthWest. You can read a little about my business and venture background on Ignition Partner's site, or read some of my more rambling thoughts on my personal blog.

While many of us run personal blog sites and publish in other media, here we will try to focus on longer form, "meaty", thoughtful pieces on issues relating to start-ups and venture capital in the NorthWest. Call it content and editorial rich. Don't come here for a calendar of events, guidance on raising money, or specific company reviews. While we may promote our investments, we will try to keep it in the context of the larger theme in which we are investing. If you're interested in more information on our portfolio companies check out each writer's firm website.

I must give credit to VentureBlog in large part for inspiration. Andrew Anker and the guys at August are great guys. While they focus on Silicon Valley and macro trends affecting start-ups, NWVentureVoice will be more locally focused on the NorthWest and our unique character.

Mostly, we'll figure it out as we go along. No idea if this is a sustainable idea or not, but we're going to give it a go. Enjoy!

Posted by Martin at 1:49 PM | Comments (0) | TrackBack