VOIP (voice over IP) is a disruptive technology that will continue to improve as it simultaneously endears itself to mainstream consumers. Over the next five years, industry experts predict that the number of VOIP subscribers will zoom by at least 40% per year. Some think it may double every year. But, revenue growth will lag subscriber growth and, as we will see below, profitably is another question entirely.

The spoken word originates as an analog sound wave and is received as an analog wave by the ear. But between these two endpoints - the audible signal and the receiving ear - the trend is to digitize everything. Once digitized, sounds can be manipulated like other digital data. In digital format, sound gets to participate in the flattening technology that should inform - perhaps more than any other - a world view of near-term technology: eXtensible Markup Language (XML). VoiceXML is to voice data as HTML is to web pages, a way to tag sound data so that it can be intelligently processed and shared. This will have profound implications for call centers, voice recognition, and yet to be discovered applications.

In the case of VOIP, a conversation is broken up into tiny packets each less than 30 milliseconds in duration. In this way, a one-minute speech is parsed into more than 2,000 tiny sound packets - each a sort of envelope holding something smaller than a bleep - that are sent over the internet and re-assembled at the receiver's end. This re-assembly is an engineering marvel because the connection seems like an ordinary landline conversation. In comparison, an ordinary connection inefficiently hordes the circuit entirely, like a hose that pumps only a trickle of water but leaves much capacity unused. In the case of VOIP, packets do get dropped due to traffic congestion but the technology keeps improving; e.g., smart network devices can prioritize voice packets ahead of data packets.

The core telephony backbone has long been digital. VOIP is one of the technologies participating in the slow, gradual, expensive up-ending of the so-called "last mile" into the home. This is the edge, where everybody from Time Warner (TWX) to Microsoft (MSFT) to Intel (INTC)to Verizon (VZ) are racing to win pole position in the consumer's mind.

But the edge concerns small business, too, and the dizzying array of edge devices: cell phones, PDAs, tiny notebooks, iPods. Here the burning question has been, will consumers want one integrated device or many specialized devices? If she wants one device, is it a phone that acts like a computer or a computer that acts like a phone? (History says that "usability" matters most, yet technology companies can't help themselves from pushing more features...)

In a nutshell, the phone companies, the cable companies, the personal computer hardware and software makers all want center stage in your living room. Everybody is talking convergence, again. Convergence has several meanings, but one inexorable convergence is the one where voice converges with packet-based data - to be carried over packet-based networks just like other data. This is not just a landline competition, however. Wireless providers are also keen to participate in VOIP. According to Annabel Dodd, "Skype, Vonage, BT Telecom, and Net2Phone are in various stages of testing or offering 802.11 handsets [note: 802.11 is the basis for Wi-Fi, the very promising short-range, broadband standard] for voice in conjunction with their broadband Internet Services. Cable TV and incumbent telephone companies are also looking at supplying Wi-Fi handsets that in the future will operate on cellular and WLAN networks." So, it's only a matter of time before the engineering experts produce a mobile phone that will be able to switch over, seamlessly, from the cellular network to VOIP-over-WiFi as you roam into a Wi-Fi hotspot.

Vonage is similar to Skype, which is now a unit of eBay (EBAY) in that neither own infrastructure: they piggyback on their customer's own broadband connections. Skype is really a peer-to-peer (P2P) technology that creates a direct, private computer-to-computer connection over the internet. Vonage, on the other hand, uses the public switched telephone network (PSTN). Skype can do this, too, but their customers pay extra for this because Skype must pay large carriers for the right to connect to the public networks. Either way, Vonage and Skype have a cheaper way to carry voice. That's the good news.

The bad news is that pure-play VOIP companies may never reach profitability. Into the model below, I plugged statistics borrowed from Vonage's recently filed draft IPO prospectus. This model attempts to illustrate the economics of a single, average subscriber:


This chart shows a timeline in 3 month intervals. Vonage spends an average of $208 to acquire the subscriber, mostly for marketing. A key question is, therefore, how long before they recoup (or breakeven) on this acquisition outlay? The bar charts at the top project quarterly gross and net profit. Gross profit is revenue minus direct product costs (in Vonage's case, average monthly revenue of almost $27 minus 48% direct product costs equals about $14). Net profit further deducts the operating costs (not including marketing), also known as overhead, which in Vonage's case is 57%. The net profit is therefore negative because 48% plus 57% equals 105%.

To answer our breakeven question, we need to look at cumulative profit. That simply means that we keep a running total: the cumulative adds the profits back to our original negative position. (You may have noticed we aren't charging interest for the cash used - aka, the capital cost - which we should. This would properly incorporate the time value of money and make the picture even worse for Vonage.)

The profits decrease because Vonage loses or "churns" 2.11% of its customers every month. Compared to wireless providers, who measure churn religiously, this is high. Wireless providers tend to fall between 1 and 2%; Verizon's is the industry leader and almost broke 1% in the last quarter. The difference in monthly churn between 1% and 2%, after you've spent so much to acquire customers, is huge.

Under the scenario implied by the S1 filing, it takes six quarters or about 18 months to breakeven on gross profit. But net profit, the number that matters, never breaks even. Therefore, in order to reach profitability, this economic profile needs to improve dramatically. And that is the whole ballgame for Vonage: they are hoping to grow into a scale large enough that, basically, the indirect costs above shrink enough to make them profitable. The direct costs, as a percentage, won't fall too much; reductions in direct cost will be met with downward pressure on retail price plans. Acquisition costs are, surprisingly, already less than several wireless carriers. Still, they could come down. But the big bet is that enough subscribers can be aquired to make the cost of running the business very small (i.e., the indirect costs). You can see how difficult this will be.

Given the valuation discipline we advise for Advisor subscribers, this is not a good investment. The problem is that, to predict profitability, your model needs to assume some future inflection point. A significant change in the business model that requires too much imagination. We love disruptive technologies, but they need to move up the value chain: we would need to see a way for VOIP providers to add services on top of the voice commodity such that the death spiral of deflating prices could be averted. We don't.

How can this be? How can VOIP be so inevitable yet pure plays will have such a difficult time? Simply because they face relentless competition from industries that have more favorable economic models. Clearly, incumbent wireline companies - who currently offer Digital Subscriber Lines (DSL) - will compete. But they are plagued by the quandry of canabilizing their own core service and, in some cases, they frankly still move like monopolies.

I would bet on the cable companies. Their advantage is that they can bundle VOIP into packages that include high-speed internet access and digital television. Comcast (CMCSA) is currently hitting on all cylinders. They added only 200,000 VOIP (called Comcast Digital Voice) in the fourth quarter, but look for them to add at least another one million in 2006. They pass over 40 million potential subscribers. Their five-year plan to add 8 million VOIP subscribers, 20% penetration, is doable. Unlike Vonage, they already own the pipe into the home and they can distribute marketing and overhead costs over internet and television offerings.

We recommend you play VOIP indirectly, by identifying companies that can profitabily leverage the technology into other offerings.