Thomas Howe

Germs, Guns, and BroadSoft

Tue, 30 Dec 2008 12:11:57 -0500

By Thomas Howe

One of my favorite books, Germs, Guns and Steel, asks a simple question: Why did Columbus sail across the ocean to meet the Indians, rather than the other way around? Why were the European and Middle Asian cultures more advanced than the native Americans? For the author, the answer came down to fundamentals of geography (the European continents were more horizontal, such that domesticated crops were more successful leading to higher populations), the numbers of native animals that could be domesticated (mutated germs from domesticated animals were the main source of human-borne illnesses, most were from Europe which increased the Europeans resistance to them) and natural elements like steel (allowing the Europeans to have advanced tools and weaponry.) As we look at the acquisition of Sylantro by Broadsoft, we can ask a similar question: Why was it that Broadsoft was the victor and Sylantro the vanquished?

Before I get too far in the analysis, I suppose some caveats are in order. First, even for people who worked directly for the two companies, this is an impossible question to answer fully. A major element of understanding organizational behavior is that of blindness: You might understand your job, but as the distance increases between your position and somebody else's in the same organization, your understanding goes precipitously down. Mike Tessler, for all of his obvious gifts, simply cannot understand what it is like to be the staff engineer at Broadsoft, or vice versa. It is nothing but hubris to believe that somebody outside either organization really understands what happened inside. Second, my own experiences color how I perceive the world. I was the US CTO of their joint major competitor, Netcentrex, which was acquired by Comverse in 2006 for $164 million. I ran the Sylantro Mashup competition in 2007; I won the Broadsoft Mashup Competition in 2008. Thus, I am as well-educated as an outsider can be, but I see things based on my own experience and am not party to all events inside the companies. Lastly, I firmly believe that analysis like this says as much about the analyzer as they do about the answers they give. However, this is valuable in-and-of-itself. So, given that I've dispensed with the disclaimers, how do I see it?

First, I think this particular segment of the market had a fishing problem. I have grown to believe that if you had to pick a single function of a business to be excellent at, it would be marketing because it answers the most important question: Does the lake that you are fishing in contain fish? From the outset, the market for IP Centrex was immature and not well-developed. Sitting here in 2008, it is much more mature and attractive, but in 2000 - not so much. IP Centrex requires a value chain that didn't exist, making it an uphill battle for all companies - Broadsoft, Sylantro and Netcentrex alike. At the time, plain ol' centrex didn't have the reputation as the hottest and best selling aspect of telephony. This set the bar pretty high on execution for all the players - failure to execute perfectly was deadly. If the pond had more fish, it would be much more forgiving. As it was, forgiveness was not in the cards.

Secondly, sticking with marketing, Broadsoft started its life with a killer marketing team, with Mike Tessler, Scott Haufpair and especially Scott Wharton. The best that any company could do was to match the skill and talent of those three; I couldn't draft a team that was better. (As good maybe, not better, and I would need some serious money to do so.) Sylantro struggled early and often with marketing and execution, moving from a PBX focus to IP Centrex, losing a little bit each time, finally settling down with a carrier focus. I met the Broadsoft team soon after they formed at a VON show, and, from the start, through the past years, to today, Broadsoft was focused.

Finally, and maybe as a direct result of its long term applied focus, Broadsoft succeeded in gaining the lion's share of the customer base, making it increasingly risky to choose another path. Even though my company had more lines and seats than Broadsoft when we were acquired, Broadsoft had more customers, reach and influence. Our customers had larger installations, with business models that lined up with our offering - but there were fewer of them. Broadsoft succeeded in establishing the standard for the industry, and is a position that they will enjoy for a while. This had the obvious impact of making it harder for Sylantro to garner attention, money and customers. The death spiral took full hold, resulting in the acquisition of Sylantro.

The lessons?
Find a pond with fish in it. Since manufacturing demand is a lot like pushing on a string, accept that it's wiser to compete for fish than to be alone in a boat on a lake with no fish.
Focus. Accept that your company may fail because you picked the wrong market, as it will allow you to win if you picked the right one.
If you aren't perceived to be the leader in your field, pick a new segment to be a leader of - even if it's smaller, because leaders like BroadSoft win.

Doing and knowing

Mon, 20 Oct 2008 10:36:57 -0400

by Thomas Howe

Today's service providers are real "doers." From business model to world view, carriers see the world in terms of what they can do for their customers. They excel at doing things like connecting two phones, sending a text message or creating a conference. Larger operators do things like write custom telephone applications or integrate enterprise software into call centers. In theory and in practice, today's service providers work off of a single business model : take a rare thing, and charge the subscriber for "doing" something with it. For years, this has been a lucrative path, and still widely is, but the tide has changed, primarily because the "doing" has gotten easier. Open source software, multiple transport paths such as cable, fiber and the myriad wireless options all contribute to an exponential rise in the number of service providers able to "do" something for businesses and consumers. This increased supply adds pricing pressure to providers, forcing them to seek better approaches and business models.

Tomorrow's service providers will rely on a very different sort of model: "knowing." Instead of generating value by "doing" things for people, they will generate value by "knowing" things that they learn by "doing" things for people. Focused on data collection, the carriers will be experts at your social network, because your real social network exists in your call detail records. They will be knowledgeable about where you spend your time, because they can track your cell phone. They will know what sorts of TV shows you watch, because they deliver video to your handset, or, because of quad play, they provide your home TV service as well. Unlike "doing", which is rather easily replicated, "knowing" is nearly impossible to replicate. Sure, a new carrier can start to collect your personal data, but they cannot recreate your past data. That sort of data is safe and sound in the walls of your current operator, and it's not going anywhere, and one day - because the data isn't moving - neither will you.

What sorts of things can a service provider know? The obvious ones are personal information and context. The trusted billing relationship between carrier and customer provides information like address, age, family structure, gender and nationality. Only the carriers can really determine our life context: where we are, who we talk to, if we are in a moving car... STL Partners identifies seven different categories of digital assets a carrier has access to, including information about relationships, interactions and devices. All of these assets are examples of things that carriers "know" - all things that are difficult to reproduce and all things are that are immensely valuable.

The problem? Privacy. At this point, you hear the privacy objections immediately. Here's what's missing from that: permission. There are all sorts of times when users gladly give up their personal information, and there's no reason to think that this is going to be different here. When consumers see giving personal information up as a benefit, and they're asked for it, they will generally give it. Location based Apple iPhone apps are a great example: when I want to use Yelp to find a place to eat, and it asks to use my current location, I'm annoyed that it even asked. I'd rather it just do that automatically. The value I find from finding a great place to eat far outweighs my concerns about who knows where I am. One day, the carriers might power the Barnes and Noble member card service. For a 10 percent discount on my purchases, I will allow the bookstore access to my data. As I walk into the store, the carrier will notify the shop keeper that I've entered, and send coupons to my phone based on my past purchases. Maybe they know my wife bought this book yesterday, or suggest one that a Facebook friend bought last week. As I walk up to the counter, my smart device will automatically pay for my purchase - no need to whip out credit card or member card. For me, the consumer, it's all good. For the carrier, it's better than good - it's the final answer to the dump pipe problem.

Thomas Howe is a long-time telecom consultant, writer and speaker who is the CEO of the Thomas Howe Company, providing expertise in improving the business process with real-time communications. His website is at http://www.thomashowe.com

Will BT Roll a Strike?

Doug Mohney — Tue, 05 Aug 2008 16:15:10 -0400

By Thomas Howe

In the book Inside the Tornado, Geoffrey Moore describes technology adoption in large markets by way of a bowling analogy: If you want to get a strike and knock all the pins down, you must first knock over the lead pin, and at least a few pins next to it. These pins take out the ones behind it, resulting in a strike.

In order for a technology to dominate a large market, it must first dominate several sub-markets, powering complete products, channels and ecosystems that make entries into other sub-markets nearly trivial. The combined effect is market domination, and once the first few pins are down, the rest almost always follow. This effect shows in every technology market, and can not only be seen in modern technologies like personal computers and spreadsheets, but in ancient technologies like calculus, basic technologies like irrigation and even footwear. It is truly a universal and a time and market invariant phenomenon.

As our market grapples with the implications of BT's recent acquisition of Ribbit, this understanding of market maturation is a critical tool for understanding the larger market. In the Telco 2.0 space, there are number of different pins to be knocked over, and there is no doubt that the first pin for BT's consideration are CEBP deployments for large enterprises. Like every large carrier, BT has long standing, strategic and lucrative relationships with domestic enterprise customers. These maturely managed companies make deep investments into enterprise software to manage their business processes, resulting in large economic savings and increased service delivery quality.

When faced with the mounting evidence of the success of CEBP deployments, they will all demand the ability to integrate their business process with their communications infrastructure. This turns out to be more than a churn avoider for BT, as CEBP deployments also represent lucrative new service offerings and professional services opportunities. For BT, providing this access into their core switching is not an "if", but a "when." Without regard to what happens outside of BT's existing customer base, they must address this need before somebody else does. As large carriers look down the alley, enterprise CEBP is the head pin.

How does Ribbit help? For all the solid business cases that surround CEBP, the fact remains that communications technology is hard technology, and the number of qualified CEBP engineers in the world would surely fit in a jumbo jet. It's not even clear there's enough talent to truly satisfy BT's potential market, never mind the customers of the other 500 world-wide operators of any size.

Ribbit's service creation environment providing Flash and Flex interfaces radically increases the amount of developers available for CEBP deployments and substantially lowers the barrier to entry to creating communications enabled applications. Even though Flash and Flex are essentially web centered technologies, and are relatively uncommon choices for the typical enterprise developer, they are easy to learn and deploy, and nothing's stopping Ribbit from providing SOAP, REST or JavaBeans on their current platform. Ribbit brings the ability to scale the implementation of CEBP deployments for BT.

Does this set BT up to attack the customers of SwissComm, Orange and Verizon?

Yes and no. It is very clear that, in relation to the enterprise, BT is not unique. Every incumbent carrier has similar domestic relationships; old, important, strategic and entangled ones. These relationships will hardly dissolve overnight, without regard to how compelling and lucrative the applications enabled by Ribbit's technology are. However, this certainly becomes a situation where the clock is now ticking, and if TelMex refuses to provide an API to El Banco de México, they have an alternative. This also represents the first, clearly identified bowling pin.

If BT is able to dominate the enterprise CEBP sub-market, it sets them up for the second pin: two sided communication service offerings. That said, I expect that BT will be have a fight on their hands to dominate this market in the medium term, as they aren't the only operator with smart guys in the strategy department ... they are the only ones with Ribbit. Ribbit's a hop ahead; BT needs to work to maintain that lead. If an operator fails to hold on to their current enterprise customers, the best they can expect is a 7-10 split.

Thomas Howe is a long-time telecom consultant, writer, and speaker who is the CEO of the Thomas Howe Company, providing expertise in improving the business process with real-time communications. His website is at http://www.thomashowe.com

Howe: A Telco 2.0 Technology Primer

Mon, 07 Jul 2008 15:09:51 -0400

by Thomas Howe

In my previous article, "Introducing Telco 2.0 - A Business Primer", I explored the fundamentals of Telco 2.0 business models and cases. In this article, I'd like to touch on the technology side of the equation, and try to define the technologies and architecture that distinguish Telco 2.0 applications from those built on more traditional technologies and approaches. The real excitement that surrounds Telco 2.0 comes from compelling business models and the promise of a final escape from per-minute charging models, but the application of web technology to voice services is the clear enabler.

A Telco 2.0 application is an application that uses other networks to carry voice, uses web services to integrate (or to be integrated) into other software services and typically has voice in a supporting role to a larger application. Let's look at these points in detail:

Over the Top: Telco 2.0 applications are almost always deployed as "over the top" solutions. A terrific example is Grand Central from Google. Grand Central is an application that accepts an inbound phone call, then rings all of the other phones in your life. Once you answer the phone, it allows you to record the call, and if you don't answer, it saves it in a visual voice mail box and emails you a link. Other than accepting the incoming phone call, Grand Central works with all of your other phones from any carrier, and here's the important part, without the carrier's permission or consent. As far as Verizon is concerned, calls from Grand Central are indistinguishable from any other phone call in the network. From a carrier perspective, this is truly scary, as the Grand Central delivers the value and Verizon becomes what they most fear: a dumb pipe. This is the first element that allows Telco 2.0 applications to scale: they ride on the backs of elephants.

Light Weight: Telco 2.0 applications are typically written with a light-weight approach, and rarely try to be all things to all people. Jott comes to mind here. Jott subscribers load their contacts into the service, then use any phone to leave messages for the people on their list. Jott takes the voice mail message, transcribes it and sends it off as an email. For the mobile in life, it provides the fastest way to send status, assign tasks and keep in touch. What else does Jott do? Not much. Like a screwdriver, all it does is its job, and it does its job very well. This is the second element that allows Telco 2.0 applications to scale: the feature sets are manageable and can be well tested.

Web Technologies: Telco 2.0 applications are based on web technologies, and nearly always integrate well with other web applications using web services. Lypp is Erik Lagerway's new startup, providing web based conferencing services. The differentiator is the API: it is drop dead simple for any web programmer to put conferencing on demand into his application. Consider how an engineer only a few years ago could enable this functionality, which now only requires a credit card and a an afternoon with a text editor. This is the third element that allows Telco 2.0 applications to scale: the Internet was built to scale.

Voice is a Spice: Telco 2.0 applications aren't about Telco--they're about anything else. Instead of the phone being the star of the show, it plays a bit part. Twitter is a shining example, as Twitter is about micro blogging and keeping in touch with those you care about. The phone is an interesting way of getting that job done. This is really apparent in communications business model applications, such as using voice messages to alert a population during a natural disaster, text messaging to vote for a game show or the thousands of pay-by-phone applications. Voice is the pawn, not the King. Voice isn't the meat, it's the spice. This is the last element that allows Telco 2.0 applications to scale: by forcing voice and messaging to be a commodity, it can permeate any other application, radically increasing the places where we can use it.

Some say there are 1,000 Telco 2.0 applications, but I would say they underestimate that number, as voice and messaging will show up in a hundred applications in a thousand niches. The technical underpinnings of Telco 2.0 make those applications inexpensive to build and deploy, and will scale to impressive heights.

Introducing Telco 2.0 - A Business Primer

Doug Mohney — Tue, 10 Jun 2008 20:17:26 -0400

By Thomas Howe

Name a Web 2.0 business model that doesn't depend on advertising. Anyone? Beuler? We may all be forgiven when we fail to name a second business model, as the success of Google, Facebook and MySpace drive both investors and customers into such a frenzy that countless startups are funded with no other value proposition than to be able to aggregate large amounts of subscribers such that one day ... we could advertise to them.

For those who struggle with naming a second model, I would like to introduce Telco 2.0. In the first of a two part series, I want to put forward the unique business proposition Telco 2.0 offers and how it differs from everything that went before it.

From a business perspective, three points are important to understanding why Telco 2.0 is such an interesting place to invest money, time and attention. Telco 2.0 refers to telecom software and services that sit "over the top" of other services such as the Internet, the PSTN and mobile networks. These applications interface with the networks below them to provide services, and use the Intranet as the way subscribers are attracted and acquired, and as the way subscribers manage their services.

As a direct result of this approach, Telco 2.0 has three identifiable business advantages over the last generation of telecom application deployments:

1) Telco 2.0 applications do not require massive investment into infrastructure, as they ride "on top of" other networks. Telco 2.0 applications connect to these networks through Web Services APIs, which are paid by the transaction and avoid investment into "big iron" approaches. A decade ago, multi-million dollar investments into hardware to support a new service was a common requirement. With voice APIs, the only requirement to support my development is a credit card with a hundred dollar credit limit.
2) Telco 2.0 applications can be successful with a small handful of customers. Since massive investment is unnecessary to develop the application, insanely large revenue targets are unnecessary. Since the service runs on top of the existing world-wide Internet and PSTN networks, everybody with a connection to the legacy networks can sign up for the Telco 2.0 service. In essence, exactly like a web site, you can deploy your application to the world with a single click, and exactly like a web site, you only need to capture a very small piece of the pie.
3) Since a small number of subscribers can make a service successful, Telco 2.0 applications can be extremely targeted, extremely narrow and extremely valuable to those who need them. The direct result of this is that Telco 2.0 applications can charge significantly more than traditional services, more than making up for whatever inefficiencies may arise from designs that ride on top of another's network.

As I meet and speak to those outside the telecom world, I often see the eyes glaze and the disdain rise when I mention "operators" and "carriers." Just like good business models, respect for traditional carriers is sometimes hard to find. My only response is simple: Name any other industry that has convinced every adult on the planet to give them $30 a month for a completely renewable resource. (In fact, isn't that the real problem? It's too renewable?) I believe that there's more than one way Telco 2.0 will flatten the traditional telecom market, and that's really going to be impressive.

Now, other than the advertising industry, name an industry so poised to take advantage of what Web 2.0 has to offer, and if you do, I'll ready my resume tonight.

My bet is that I'm going to bed early.

Howe: Congratulate Failure

Wed, 28 May 2008 16:01:22 -0400

By Thomas Howe

The recent demise of Jangl is an excellent sign for the overall health of the Telco 2.0 marketplace, and although difficult for their investors and employees, portends great things for this market. My sincere wish is that Telco 2.0 failures become more commonplace and numerous, and I firmly believe it will be so. As stakeholders in this market, we need to expect and encourage such failures. Congratulations to those involved, and I mean that sincerely and with no animosity whatsoever.

Not too long ago, new telecom services and applications were costly to develop and deploy, with only a chance of success in the end. Even for those that succeeded, competitors typically replicate the service in short order, gutting profits. The logical conclusion of many managers is to be cautious with new service deployments, and to concentrate on maximizing existing services, stifling innovation. They are afraid to fail, because failure costs money. However, something's changed.

What's changed is that costs of service deployments are falling at an increasing rate. The Internet provides a near frictionless marketplace for both service delivery and sales, and the challenge is no longer the design and development. The trick is now fundamental marketing: what do people want? How do I price this? What is the fundamental problem I can solve? Surely this is no small task, as these questions are notoriously difficult to answer and are as often discovered through dumb luck as solved through excellent analysis. There's no way the existing telecom infrastructure could support the development of a million applications, with the hope that ten thousand will be successful. However, a Telco 2.0 infrastructure can and will.

Telco 2.0 applications are built using web technologies to overlay functionality on the existing PSTN infrastructure. Since the PSTN infrastructure exists, there's no reason for the Telco 2.0 engineer to build the capacity for switching calls around the network. Since Telco 2.0 applications are delivered through the browser, there's no large workforce to manage and train. Since there's no natural geographic barriers, the entire world becomes your serviceable market. Since web technologies are naturally scalable, and are paid on a transaction transaction basis, large capital investments are unnecessary. As a result, it's radically simpler and cheaper to write and deploy them as compared to traditional services.

The result is that we'll see many more Telco 2.0 applications than we've ever seen from any other telecom market. Ever. Nearly all of them will fail, as most of the people developing them will not have the marketing skills to have any real chance of success. They tell authors that about one in a thousand books are actually published, and the number of authors that have published more than one book is smaller still. As you enter your local Barnes and Noble, start counting the books, and realize each represents a thousand others you'll never read. However, without mechanisms that allow people to write books simply and cheaply, how many of the books that fill the shelves would have been written in the first place? I say not many.

Congratulations to all those authors of software and prose who remain anonymous. Congratulations to my brothers and sisters at Jangl. Even though success was not personal for them, in very real and important ways, they guaranteed the success of their markets.