Saturday, October 18, 2003

SMART or Stupid?

Ok I admit it, David Isen is smarter than I am.

[non-sequitur] But that being the case; if David is right and it's better to have a stupid network than an Intelligent Network, maybe it's better to have a stupid network pundit, too? [/non-sequitur]

So the telcos are, according to David, doomed. Either they must irrevocably alter their business strategies, their thinking, and their practices, or they will die.

Either way, it seems, they will never again be what they were.

I know, some of you are still waiting for the downside. Don't worry, there isn't one. Not for you anyway. But for telco executives weaned on the presumption of market monopoly power and the concommittant irrelevance of pricing or efficiency to profits? For them, it seems, the bell tolls.

If you haven't been to isen.com and signed up for David's SMART Letter, then I recommend you do so. You won't always agree with him - hardly any point in reading him if you do - but he always has something to say worth disagreeing with.

In my last missive, I proposed that the problem the telcos face relates directly to the growth of transport capacity to the point where it is now "cheap as dirt" and at a pace that far outstrips the growth in avaliable capacity to make intelligent decisions about the traffic riding that transport.

It's like what happens at your local border crossing, where 5 guys are trying to make intelligent choices about who to let across the border. They have to slow each vehicle to a stop, closely (i.e. slowly) inspect its contents, particularly the people, think about whether or not those contents are legitimate cargo, etc.

This all takes time, as anyone who has ever waited at the back of the resulting traffic jam knows.

But the problem of people checking on other people is directly analogous to microchips checking on microchips: there is no single microchip that can effectively audit the output of 1,000 others of the same type.

Yet for some perverse reason of their own, the telcos are still vainly attempting this insanely hubristic feat.

There are even telco equipment manufacturers out there who are prepared to take the telcos money while promising them exactly that sort of "solution" to their problems.

And while I hesitate to call them snake oil salesmen, because they have more lawyers than I do, nevertheless the urge is there.

Copyright (C) 2003 Denver Fletcher. All Rights Reserved.

Share & Enjoy

Tuesday, October 07, 2003

Smart Network is Oxymoronic

Hmmm ... ideas.

For me, blogging is a way to get ideas out of my head and into the "Real World"(TM). It's a way to look at them a bit more objectively, to examine them maybe a little more dispassionately, to get some feedback (hopefully) on them, to tune them, refine them, assess them, discard them, classify them. It's an artistic practice of what should perhaps be a science. In short, more a hobby than anything else.

I have a personal motto I use in my email .sig file: "I have no monopoly on good ideas."

I know, it's contrary to the warholistic narcissism of our age. I'll never get to be Robbie Williams by thus keeping my ego in check. (Of course, the fact that he can sing and I can't has nothing to do with it.)

My 15 minutes of fame has likely already been consumed by someone who got half an hour, maybe more. Probably by the same bum that got my share of vocal artistry.

Warren Zevon - or was it Frank Zappa? someone with a Zedname anyway - once said that he'd had his share of mind altering substances; his and two or three other people's.
(It was Jackson Browne, - Ed.) Apparently no-one ever thought to ask how we could all get our 15 minutes of fame if celebrities keep on taking up days and days of the stuff and politicians extort iniquitous sums of money from us all so they can employ people to make sure they get weeks of it at a time. Don't get me started on what happens when celebrities become politicians!

I don't blog for fame. You won't find any blog-rolls here. No incestuous back-scratching, no mutual admiration society. I'll link to relevant websites and articles, maybe, if I feel like it, be they blog or no, because it supports, enhances, enriches, counters, or just plain fits, whatever happens to be on my mind when the blogpulse strikes me. If no-one ever links to this blog, I will still die happy, I promise you. If no-one ever reads it other than me, it will have still served its purpose.

Speaking of incestuousness, have any of you bloggerati out there considered what the natural results of an incestuous blogger "family" might be? I know you like to refer to yourselves on occasion in such terms: family, society, community, etc. All terms that tend to signify some commonality amongst you, but also that identify you as separated both by inclination and habit from the set of non-bloggers.

Being a necessarily finite subset of humanity and its glorious dirty confusion, have you thought about what the purist inbreeding you're doing in your blogs is going to produce in the next generation? Or the one after that? All furiously linking to each other, madly circling the HTML wagons against marauding print media critics and other pagan savages, cross-pollinating like crazy, building infection vectors for some hideous idea virus as yet unseen but nonetheless implacably slouching toward us waiting to be born?

(Hey now, don't get me wrong; the print media have their own insanities to worry about. Or not, but at least they worry me.)

Irrelevance would seem to be the least of your worries.

That's the problem with the whole memetic obsession with ideas as viruses; it speaks (or fails to) of the presumption of the infection process as an inherent goodness in its own right, never stopping to consider what else might ride the pony it comes into town on. For my money, I'd bet that even black knights may ride white horses.

Just an idea ...

... and ideas are funny things in their own right. This diatribe started out with an intention to write something about routing, about decision making capacity in the routers that underpin the internet, and the growing disparities between that capacity and the available transport capacity, and what those disparities might portend for the future of the internet and the priesthood of the "IP everywhere" religions.

I thought it might be amusing as an aside to make a conscious parallel between that and my rather more concise explanation of myself and my blog back on page one. I blog to get ideas out of my head and out front where I can look at them because they take up space (of which there is plenty to spare in there) and they take up thinking capacity (of which there is a terrible and chronic shortage).

But I haven't even got started on the main event yet and already I've slagged Andy Warhol, Robbie Williams, and called bloggers everywhere "motherf**kers" - in a blog of all places! - Chris Locke will no doubt uniquely appreciate the Freudian implications of that epithet.

Where will it end?

Inner Space, our own final frontier. I thought I might boldly go where no man has gone before and peek under the covers of my own head, my own thoughts, my own ideas, my own personality. Yuk! Sounds like some Beeblebroxian nightmare, doesn't it? Zaphod always held that introspection was a huge mistake.

But space, in this case, is analogous to transport capacity. It is the carrying capacity of my own mind, the ability simply to hold an idea. It is not indicative of and nor does it necessarily imply any innate ability to do useful things with or to that idea.

When I first started out in the networking subsector of the IT industry, transport capacity was incredibly scarce and correspondingly expensive. Almost unimaginably so for those now entering into our time of plenty. And what plenty! 10 Gigabit Ethernet is an outrageous amount of transport capacity. Compare this to the 50 baud I started out with, back in the dark ages of nineteen-eighty-mumble, and the then incredible riches of 1200 baud and the promised nirvana of 2400bps. And you try and tell that to young people today, and they won't even understand you, let alone believe you. To bandy about Kilo bits per second with sufficient abandon to stop counting every single bit and start referring to them using terms like 9K6 is itself a sign of increasing plenty but also of increasing presumption.

Relatively primitive and infantile though the nascent large scale integrated circuit industry was in those days, it had one big advantage over the telecommunications industry: it was largely untouched by governmental hands. It was - largely - neither regulated nor owned by governments, and so it leaped ahead in vigorous bounds, rapidly catching and surpassing the capacity of the hidebound government monopoly telecommunications carriers almost as soon as it was born.

As PC's became fashionable (not that anyone ever admitted to mere fashion being a driving force in business buying decision making, but don't let them fool ya, oh no!) we started looking at various ways to join them together. One of the first I came across was DECnet. Horrible thing. But it ultimately rode on Ethernet cable, which wasn't all bad.

It was bad enough, back then. It was two inches thick (it is still called "thicknet"), it was as unbending and unaccommodating as your wife at 3AM after your return from a night out with the boys. You had to measure out so many inches and carefully tap into it to get a working connection. If you did it wrong, all sorts of unpredictable results could happen, but the most common one was that nothing worked and the whole effort was a write-off. Hmmm, maybe my wife analogy isn't too far wrong at that. It was as expensive as any wedding, and I'm sure you know that the consequences of screwing that up are not good.

But eventually a number of trends caught up with the telco industry and forever changed it. Increasing deregulation infected the industry globally. As more governments withdrew their overbearing oversight, competitive innovation was allowed to show its hand, especially at the transport layer. Latent demand for capacity was already huge, having built up a massive head over decades of stonewalling and sandbagging; with fibre-optic technologies coming into their own, the supply exploded out to meet this demand as soon as it was free to do so.

Transport capacity is now incredibly cheap and increasingly moreso. The monopolists still cling to their "last mile" copper monopoly by dint of the economics of low density residential housing, but in high density and/or high value districts, the monopoly party is long over. Competitive transport suppliers have put so much potential capacity in the ground that the sum is almost incomprehensible. In direct relation to this burgeoning supply, the per bit price of data transport has fallen off the cliff.

The problem with thus being able to carry a stream of data packets at such incredibly high speeds is that the decisions about where it needs to go cannot be made at the same speed.

It's like this: the same clock cycles, built on the same underlying timing mechanisms, control both processes:
- the timing of voltage or photon emission transitions that make up binary digital signals, and
- the timing of control path decisions about where those electromagnetic or photonic impulses get routed to.

But (there is always a "but", eh?) decision making takes a lot more clock cycles than simple bit boundary transitions do. So using the same fundamental timing mechanism to drive both processes is always going to leave an intelligence deficit in the system, right?

Now those of you who've ever had any exposure to telecommunications companies can better understand their behaviour! (That's part of my anti-telco campaign. They're not all bad, really. Just stupid. Heh.)

It is a lot like Einsteins famous quote about how the problems we've created cannot be solved by resort to the same level of thinking that created them. You can't route packets at the same rate you can make them.

It's getting worse. The CPU manufacturers had a head-start on the telco equipment manufacturers. That head-start has long since been eliminated and reversed. Today, you can put 40 to 80 Giga bits per second on a single wavelength and one fibre strand can carry 10's to 100's of such wavelengths.

By way of contrast, the CPU manufacturers are struggling to put 2 or 3 GHz CPU's into your laptop because the heat production is already killing them. They can do it because the CPU's typically operate at an avergae of 4% capacity. If you want to really test the life expectancy of your 2.4GHz laptop, then download one of the cooperative prime number solving programs and let it run your CPU at 100% for a while. Don't do this unless you can afford a new laptop though.

Will they be able to put 10GHz in there? 20? 40? 80? 40 times 80 perhaps? Seems unlikely, doesn't it?

Moore's Law says that chip capacity (roughly equivalent to processing power, which in turn implies decision making capacity) doubles every eighteen months. But unfortunately for the network intelligence deficit, transport capacity is doubling every NINE months! The disparity is not only growing but accelerating! The yawning chasm deepens.

What is the answer?

IS there an answer?

Well, there's always answers. Which answer is best I leave to another day. But go and see David Isens 'Rise of the Stupid Network' for a vastly superior explication of both problem and potential answer than I am capable of.

Copyright (C) 2003 Denver Fletcher. All Rights Reserved.

Share & Enjoy

Friday, October 03, 2003

Mobile Messaging

Your average telecommunications company executive is probably a pretty smart person. He (usually it is a he, very occasionally a she) probably has an engineering degree, often enough an MBA or a marketing degree as well. He probably has an IQ over 100, but maybe not much.

He typically runs a company that own thousands of miles of copper in the ground, and leverages this natural near-monopoly to obtain near-monopoly profits.

But the success of this strategy (that's a bit flattering really, since it wasn't planned at all, it just happened by some accidents of history for which the present management can take no blame nor credit) is also the greatest danger to telco companies, and thus to telco executives. It has bred a distinctly cynical apathy into them that is leaving the whole industry increasingly useless and irrelevant.

"What does this have to do with mobile messaging?" - I hear you cry.

Well, the thing is, the mobile or cellular phone service companies are often owned by the old wireline telco companies. Sure, there are pure-play mobile companies in increasing numbers, and that's a good thing. But in every market you will almost invariably find that the old wireline monopoly owns a significant chunk of the mobile market as well, and quite often owns the market leader outright.

There is some sense in this; they share many of the same customers, and those customers want the convenience and the economic efficiency of a single supplier relationship for all their telecommunications needs.

But unfortunately, many of the mobile executives of today were born and bred in the old wireline businesses of yesteryear. Even in the pureplay mobile operators a large number of this breed are entrenched. Most people in the industry assume (perhaps unconsciously) that selling mobile telco services isn't that different to selling wireline telco services.

But, in fact, it is radically different. Because what is being sold is actually mobility. And mobility is the direct opposite of the stasis that is most strongly implied by fixed copper lines buried in the ground; a stasis that somehow also seems to have long ago invaded the very thought processes and genetic makeup of the vast majority of telco executives.

Now, you may think that I don't like telco executives very much, you may think that I enjoy publicly excoriating them far too much for it to be healthy. You may be right. But that's not the point of this rant. Rather than simply lambasting them for their habitual lack of creative thought, of insight into their customers needs and wants, and so forth, I thought I'd help them out by way of a simple illustration - Mobile Messaging.

Mobile Messaging includes SMS, paging, some CDPD applications, some WAP applications, and similar uses of newer mobile data transport technologies like CDMA-1X-EV-DO.

I'm going to talk mostly about SMS because it's simple, it's increasingly ubiquitous, it has a respectable customer base, and it is seriously under-appreciated and thus under-exploited by every mobile operator I know of. By extension, that also means that it is seriously misunderstood, mis-under-appreciated and mis-under-exploited by every wireline telco on the planet, as well.

But mostly because most of you will be familiar enough with it for this rave to hold your interest. Because who wants to listen to some lunatic raving about something you neither know nor care about?

Metcalfe proposed that the value of a network is proportional to the number of customers (connected nodes) multiplied by the revenue per customer. He further postulated that the revenue per customer is proportional to the number of customers; in other words, the more people who have a telephone, the more they will talk to other people on the phone, and thus the more revenue that each customer will generate for the service provider. This seems not only to make sense, but to actually be true in practice.

Thus, Metcalfes Law says that the value of a network (Vn) is proportional to the square of the number of customers (C2).

We can express that Law like this: Vn ~ C2

If you look at this equation in its more primitive form, you might express it like this: Vn ~ C * Rc

Where Vn= Value of the network, C=number of Customers, and Rc=Revenue per customer.

Such an expression might lead you to believe that there are two major levers that a network operator might have available to increase the value of their network: increasing the number of customers, and increasing the average revenue per customer.

And indeed, if you examine every single mobile operator on the planet, and all the many pundits, commentators, and other parasites that collect around them, you find an overweening focus on what they call ARPU (Average Revenue Per User) and this metric is commonly used to measure the success of that operator, and in turn the desirability of its stock. It is not uncommon for the announcement of a significant growth in ARPU to be immediately reflected in an increased demand for that operators stock, and thus a higher price attached to their stock.

This obsession, as we shall see, makes absolutely no sense, but that doesn’t stop it from occurring.

There is nothing particularly wrong or wrong-headed with investing in new services that entice your existing customer base to spend more with you than they do now.

But focussing on revenue per customer to the exclusion of number of connected customers is madness. Look – which would you rather have: a network of ten customers with ARPU of $100 or a network of a thousand customers with ARPU of $75?

Not a trick question.

Also, as Metcalfe said, if you increase your number of connected customers then the revenue per customer tends to increase proportionately anyway. But increasing your revenue per customer doesn’t necessarily do anything to increase the number of customers. It may decrease the number of customers because your revenue is necessarily their cost.

There’s more: many mobile operators resisted the joining of their SMS networks early on, and the more popular SMS became the more they resisted, because they thought that if SMS is so popular it must be a strategic hold on the customer, i.e. they won’t want to change once we’ve captured them, we don’t want to make it easy for someone with 10 friends on our network to join the other network(s).

Do the math:
- A network of 10 customers has a value of 100 (10 * 10)
- Two disconnected networks of 10 customers on each have a combined value of 200 (10 * 10 + 10 * 10)
- Two connected networks of 10 customers each have a value of 400 (20 * 20)

Divide the 400 into two operators and each one has doubled the value of their network just by connecting it to the other.

This isn’t rocket science, is it?

The idiocy of some of these people becomes more apparent when you consider that so many of them have spent decades in the wireline side of the telco industry. Would they have ever contemplated cutting their telephone networks off from the rest of the world so as to hold on to their customers? No, they’d have quite rightly had you committed to the nearest asylum for even thinking such a thought. But they felt quite confident that with SMS it was the right thing to do. Why? Nobody knows. Today, none of them will admit to it. They don’t want to talk about it. They don’t want to hear about it. It’s the British Foreign Office: Never Apologise, Never Explain. Just ignore it until it goes away.

Unfortunately, for them, SMS isn’t going away. Mobile Messaging generically isn’t going away. It’s becoming a bigger and bigger part of their business, and of their customers lives and businesses, and they had better come to grips with it or they are going away.

OK, so now what?

How do mobile operators expand the customer base of their SMS networks?

Let’s look at some basic economics, shall we? Is there an elasticity to the demand curve for SMS service? If we made it cheaper would more people buy it? Probably.

“But we don’t want to cut our prices!”

When did you ever? Shall we play the eye-rolling game now, or later?

But look, the cost per message isn’t the only economic hurdle to becoming an SMS customer, is it? There’s the cost of a handset that comes complete with a multitude of voice functions that the price-conscious SMS customer doesn’t want and doesn’t use. That cost often represents thousands of messages.

But a dedicated SMS device could be built for about US$30. Every teenager in the western world could not only afford that but would gladly pay it. Do mobile network operators want to have every teenager in the western world as their customers? You could have fooled me!

The device should be purpose designed and built for text manipulation and display, without the cost of voice functions or the compromised controls necessitated by catering to voice functions.

With high volume low cost production underway, a multi-network high end model could replace the paging service (save the operator money!), and become a two way paging service over SMS. Global roaming could be achieved for those high end customers that want it, because it would be cheap to implement multi-network interfaces in such a relatively simple device. GSM, D-AMPS, CDMA-1X, and 3G, all in one.

Even voice customers would be attracted to such an option because the cost and complexity of global voice units is not worth the bother but the cost and complexity of a globally usable SMS terminal would still be a fraction of the typical single network voice handset aimed at the business customer today. They would find a global roaming capability for SMS to be a worthy stop-gap to the patchy cover available for any single voice service network.

What about all those wireline customers that are also customers of the monopoly wireline telcos mobile division? Does the telco make any effort to supply them with wireline SMS terminals? At work? At home? Does it make it easy for them to use web interfaces to SMS services? Does it make it easy for them to run their own info-services over SMS? Does it make it easy for them to experiment with new ideas, new services, new uses of the SMS network?

In other words: Are they actively trying to increase the number of users? (What Peter F Drucker would have told them: create a customer!)

Like hell they are!

Do the pundits fare any better? Not really.

Here’s an excerpt from a recent report published by a supposed expert:

The Register: "Swain said that the popularity of data services tends to grow with the launch of 2.5G (GPRS) networks, which often come with new applications for all types of subscribers. 'Not only will business use of wireless data for enterprise applications increase, but consumers also will find new sources of entertainment and information,' Swain said. In Ireland, O2 and Vodafone launched their GPRS services in early 2002. "

See? The consumers will find new sources of revenue for the operators. Really?

Does that not seem at all backwards to anyone? Would Peter F Drucker ever have proposed that the consumers job is to find new sources of revenue for the supplier?

Oh, sure, the consumer will find new services; they’ll find them on a rival network. They won’t long provide revenues to operators who expect them to do all the work as well as pay all the money. That’s just not how the game is played. Not on this planet.

And given the well established control freak character of network operators, who’ve been busily ring-fencing their services and trying to own all the content as well as the delivery vehicle, does anyone really think that consumers will “find” compelling (you know, enough for them to part with dosh in exchange) new sources of entertainment and information on such networks?

Does anyone remember how often the telcos have attempted to become content providers and own the whole vertical?

Does anyone have a spare amputated stump so we can count on its remaining fingers how often that has worked?

Copyright © 2003 Denver Fletcher. All rights reserved.

Share & Enjoy

Wednesday, October 01, 2003

Fugue in A Morrie Minor

A new blog ..

Why do we do it?

Who knows? Who cares? Who gives a rip?

I can only answer for myself. Possibly. I claim full responsibility for my own ravings. Indeed I claim all credit for them, real or imagined, too. Any of you sad-sacks try and pinch my stuff and - well, I warn you, I can be pretty vicious in print, so just don't even go there. The collateral damage to the reputation of anyone stupid enough to plagiarise me would be poetic justice on a truly cosmic scale.

My head is full of ... ... well, ideas. Yeah, yeah, it's been said I'm full of the other stuff, too.

I quote other people on occasion. I've developed the conviction that all of the collected wisdom of mankind can be discovered in the lyrics of our popular songs, one-liners from books, movies, TV shows, etc. Blogs now, too, I spose ...

You should understand right at the start, this is not a compliment to the various popular arts, not in any sense. It is rather a deliberate damning by the faintest possible praise of what passes for wisdom amongst the membership of our benighted species.

You'll find this a recurring theme, I think. I promise nothing.

I'm sort of guilty by association. And my despite for people who plagiarise my work doesn't prevent me taking what I can use from anyone else, within the lacy constraints of fair use exemption clauses. I make no claims to moral superiority, rational consistency, nor any of the common virtues typically ascribed to civilised men.

What was I saying ... ... ... ? Oh yeah, ideas.

Well, Ideas As Opiates is the name of a song by Tears For Fears, a badly under-rated 80's teen angst duo out of the UK.

I use it here because blogging these ideas gets them out of my head, where they consume space (of which there is plenty to spare in there) and thinking capacity (of which there is a terrible and chronic shortage).

Hence - Ideas As Opiates. Allegorical Attitudinal Adjustment. Keyboard Stress Relief. ASCII Therapy.


There won't be any blog rolls here, either. None of that damned hero-worshipping nonsense. Winer? Searls? Kiss my ...

RageBoy is the only blog I read regularly, but only because I'd pay money to read anything that lunatic Chris Locke writes, and RageBoy is free.

Later, I'll make some risible attempt to discuss some things I actually know something about.

You just wait and see.

Copyright (C) 2003 Denver Fletcher. All rights reserved.

Share & Enjoy

This page is powered by Blogger. Isn't yours?