Monday, 23 July 2007

Divergent needs

Despite its name, this blog is not about rebuilding telecommunications hardware. Its about the business of making calls and messaging more ubiquitous via software - the cables and wires can stay where they are.

Have a look at the Economist article "Home truths about telecoms". The use of communication technology is not converging. Most of us use a mixture of channels; the good old land line as the household or work "collective channel" for quasi public calls, mobile for last minute co-ordination, texting for intimacy, email for admin, IM for continuous communication.

This tells us that people want lots of different solutions, to fit different needs and circumstances. Mixing two different high technologies e.g. video + phone = videophone without working out what the need is doesn't work. Nobody wanted a videophone because it didn't fit a communication niche. Find the niche, then make the tech work.

So it should be quite possible to hit the right untapped communication niche with the right combination of software, and the right web services. A simple idea like sending a message to a different device depending on the time of day and what the recipient is doing hasn't really been licked yet. There are probably a lot of ways to mash up comms; some of which will make a strong business case.

Wednesday, 18 July 2007

Why Price Isn't Enough

Earlier today I learned that SunRocket, the VOIP only Telco had shut up shop and left 200,000 odd subscribers in the dark without their communications. With most of us now having multiple ways to communicate through our landlines, mobiles and softphones, losing one type of communication isn't probably a big problem, but if you only had one form of voice communications and that gets taken away, that could be quite painful.

Anyhoo, does the departure of SunRocket provide more opportunities to others to take the market, or is it a signal either the market isn't ready. Is it simply the case that SunRocket plans were foiled and they didn't have a plan B? It's hard to say from the outside, but there's a lesson for us all in there. Know your customer, understand the business problem, don't over commit and think about competition.

Check out these blogs for more insights.

So farewell then... SunRocket

Monday, 9 July 2007

What is a Phone Call? Pt 1 - Originate, Interconnect and Terminate

Following on from my last post, What is A Telco, I thought I'd share some of my thoughts about what a phone call is.

Most users have a fairly simple explanation of what a phone call is. They pick up the phone, and dial (or find an entry in the contact list). Their phone places a call, and the guy at the other end either answers, or goes to voicemail, or hits an error state. The errors could include "Number not available" "Cellphone out of range" etc.

But when you start digging under the covers that's a pretty simplistic explanation. It's a bit like asking a checkout girl how the barcode scanner works - "I lift up your item, scan it, and the screen says what it is and how much it costs." Course it does. Ask someone from NCR, they might have a different answer.

I'm gonna peek under the covers. See what's ticking.

Let's start by thinking of a call from one cellphone to another, in different countries, where the phones are owned by different operators, let's call them Giant Telco, and Big Cell. Arthur gets his cell phone and mobile calling plan from Giant Telco. Brenda gets hers from Big Cell.

Arthur places a call from his phone to Brenda. Except of course he can't call Brenda's phone direct, as there is no direct network connection between the two. So in fact, Arthur is placing a call to Giant Telco, asking them to hook up his phone to Brenda's. Arthur has no idea what network Brenda is on, but he does know that she's in a different country, because the dialling code tells him that. So he won't get billing surprises later. But that's another post!

Giant Telco now has this call request come in. Let's call this a call leg. It now has to find Brenda's phone and make it ring, while keeping Arthur's phone connected. Fortunately numbering ranges are fairly well understood between Giant Telco and Big Cell, so this is easy. Unfortunately Giant Telco and Big Cell don't have networks that link together. So Giant Telco needs to route the call with another carrier, let's call them Deep Sea Net, as they specialize in underwater fibre. Deep Sea Net gets the network between Giant Telco and Big Cell set up and tickety boo, then Big Cell starts hunting round for Brenda's cell phone. Fortunately it keeps a good track of where Brenda's phone is using a piece of kit called an HLR, so it knows exactly where to send the call. Finally Brenda's phone rings, she answers, completing a new call leg. Her leg is joined with Arthurs, they can talk. It's good to talk.

This all sounds expensive, and that's because the infrastructure to support all of this IS expensive. If you want to start your own telco, it's probably not a good idea to start building all of this, even assuming that you have permission to dig up roads to lay copper, get permission to broadcast radio frequencies etc. If you want to do that, then you are reading the wrong blog. This blog is about starting your own telco with none of that.

How do all these telcos get paid? Probably the paying customer in the example above is Arthur. He gives some money to Giant Telco, probably based on the number of seconds or minutes that his call to Brenda was running. Lucky old Giant Telco gets to keep the lions share of this money, but not all. The payment they get is for call origination charge, as the call started on their network, with Arthur's device being an endpoint on that network.
Deep Sea Net gets paid by Giant Telco, because the call went over their network. This is for interconnect charges. In some scenarios there may be many parties involved in interconnects, and indeed some Telcos offer various interconnect services for sale. I'm guessing there are some financial gurus and regulatory types who spend their time battling over interconnect fees. But going down that route isn't going to get you up and running very fast.
Finally, Big Cell gets paid by Giant Telco too, for terminating the call. In some cases (this happens a lot in the USA) Brenda gets charged for receiving a call. This doesn't seem fair to me, but Big Cell keeps signing up more consumers, so I guess the model works.

Similar things happen with Internet Charges. I buy broadband from my ISP, who manages their own network (though it may be a virtual network too!), and they then buy back bone from large carriers (yep, most of them are Telcos). This is another digression.

Back to Arthur and Brenda's call. I've described the experience of making a phone call that we are all used to now, not many of us really remember when we had to talk to an operator for each call we placed! Fans of old movies see this all the time. 
However there is a new breed of call, in which a central intermediary originates both calls from their own network. JaJah is doing this, it's also the way the BT Web21C SDK works. We know this as a Third Party Call. Here's how it might be implemented, Arthur presses a button on a web page, lets call it that says "Call Brenda." The web site prompts him for his phone number. Then the site calls him - it's originating the call. When he answers the site calls Brenda - it's originating the second call. When Brenda answers the two calls are connected. Arthur could've been viewing a web page rendered on his cell phone of course!

Three very special things are happening here.

Firstly, both Big Cell and Giant Telco are only getting termination fees, for terminating their respective legs of the call. It's probable (but not certain) that Deep Sea Net is still in on the game, and due some fees. This depends on the relationship has with Big Cell and Giant Telco; or more likely the relationship that's telephony provider has with Big Cell and Giant Telco. You could speculate that the bigger the provider, the better and more efficiently they can manage their interconnects - potentially on just their own network if they are large enough. This leads into the algorithmically complex area of Least Cost Routing. This is simplified somewhat these days by the internet, where a bunch of voice traffic now can go over managed IP links such as VPN's, such as might link two Enterprise IPBX's

Secondly, the billing pattern has changed. Even though Arthur clicked the button, the _customer_ requesting the call is The telephony provider will charge them. then hasthe choice as to whether they want to absorb the cost, or somehow bill Arthur or Brenda. How they do that is now up to them. Depends on their business model (ad supported telephony anyone?)

The other special thing that happened (and you may not have noticed, I didn't make it clear) is that Brenda didn't give out her phone number to Arthur. She's chosen to hide it behind the "Call Brenda" button on her web site.

Clearly the customer experience for Arthur is not as good as a cellphone originated call. But if the cost is orders of magnitude less, then it's likely that Arthur will accept having to answer his phone in order to call Brenda!

Indeed, I personally believe that this experience is better than a cellphone originated voip call, using a client such as Skype. Arthur would still be picking up costs for his data connection in that case anyway, and this normally isn't cheap. Plus the carriers have the ability to throttle voip traffic over their cellular networks if they wish. The third party call I describe here is using the plain old telephone service(POTS), that has cost billions and handles huge number of calls every day. With a quality of service we are used to, often 5 nines or more

Saturday, 7 July 2007

You and Telco 2.0

So you want to make phones rings from your website? And make some money? Well, someone will. May as well be you.

The great thing about "web 2.0" is that you can mix bits up and get something new. You just pick the best components for the job and trust to community standards to link them together. The definition of "best", however, is not so simple to figure.

Some platforms or standards will tend to dominate, or move you to conclusions that aren't quite right. And you may have prejudices that you need to get past. Much as I think Web21c is great, it maybe that another vendor is more for you. When I first read the hype about Facebook as a thriving new platform, I sort of thought it was only for kids. Then I found that half my friends were already on it.

Your job is probably to focus on what you think your customer or audience are interested in. But against that, you have to invest time in the technologies and platforms that will help you now and later. I guess Apple are a good example. They haven't dwelt on their commercial failures (remember Lisa?, or Next?) Their success is in mixing what they do well with whatever good ideas they need. The iPod was neither first player, nor following any obvious trend. And it wasn't an in house solution.

Its not as if people don't have phones already, but ubiquitous communication is coming real soon now. So find the right niche and JFDI.

Friday, 6 July 2007

Where do you start?

I needed some hosting, a domain name, a paypal account and access to a service to fulfil the call requests. I got the following set up in a few hours...

Domain: (TELCO with a '3' replacing the 'E') - £25 per year

Hosting: - £125 per year with no arbitrary bandwidth restrictions

PayPal: (feel free to send me money) - about 23p for a £10 customer account topup transaction

Fulfilment: (BT's Web21C SDK) - about 20p per call

I wanted to promote feelings of simplicity and friendliness so I chose modeling clay to create my logo and some claymation elements for the site.

Something to Sell

To make a sale I needed a product and so I chose to start with voice calls.

I can initiate a call between two phones anywhere in the world using the BT Web21C SDK at a cost to me of around 20p where the call can last for up to 2 hours. While this may not be cost effective for my customers making a quick local call to say they are on our way home, it does offer some real value for those making long overseas calls particularly if they want to involve mobile phones at one or both ends.

To make a profit though I need to add a margin that will cover other costs such as hosting, transaction fees etc... So a charge of 30p per call can still offer value over a two hour call of around 0.25p per minute to anywhere in the world.


What is a Telco?


'Telco' according to Wikipedia is a generic term for a telephone company.

I guess that means someone who sells, or gives away telephone calls.

Used to be that a Telco was a government agency, and the agency owned a bunch of copper in the ground, some buildings with switches and operators in (the exchange), and some billing infrastructure - probably this used to be people with abacuses. Hell, the agency even owned the telephone in your house. They owned the lot of it, from end to end. You got a phone from the government agency, or you didn't get one.

Things have changed a bit.

We've had decentralization. We have monopoly commissions. We've got a stack load of regulation. We have the internet. I don't think Alexander Graham Bell would recognise the telco of today. Voice data is now broken down into packets, and switched. Who knows if even one word goes over the same route?

So I think that the companies who were the Telco of yesteryear has morphed. Now they are a carrier. They shunt bits. Sometimes these bits are voice. Sometimes they are data.

Clever Telcos have extended their businesses. They are often now ISPs. They include among their customers other ISPs. They sell hosting. They sell IVRs. The one thing they seem to have in common is a whole bunch of sunk cost in network, and infrastructure. Sadly the end user experience hasn't changed very much. After all, if you pick up the phone and dial, you get through. What's different?

Well, we think that you don't need to own any network, or infrastructure in order to be a Telco. We think that anyone can be a Telco, and sell telephone calls, and advanced calling services.

This blog is going to show how you can build a telco with minimal upfront cost, on a transactional basis, and, hopefully, turn a profit!!

We're not going to tell you how to compete with AT&T, Vodafone, T-Mobile, Google, Skype, or any other Telco. How you do that will be up to you. You'll need to decide exactly what your positioning is.

All you'll need to get started is a $1000 PC (or mac!), a development environment, and some tools that we'll show you over the course of time. At some point you might need a web server. We're going to base a lot of the telecom integration on BT's Web21C SDK, which exposes their global network to developers with a very interesting pricing model.

So stay tuned, and you might find out what we mean by 'Leg and Mux'. The difference between call origination, interconnect and termination fees. What the difference between an IVR and an ACD is. The problems with voip. And anything else which springs to mind! I look forward to watching this conversation develop.