Wednesday, August 13, 2003

Telirati Newsletter #5

In this, my fifth newsletter, I take aim at a fat target: The New PC. In this case it was the "Simply Interactive PC." Remember the "Simply Interactive PC?" Nah, didn't think so. But it did suck a lot of mind-share and ink in it's day. Many lessons for today, as well...

Telirati Newsletter #5: Playing chess

I can swing a golf club. I don't know how to play golf. Just because you know the rules of chess, does not mean you can play chess. Newsletter #4 was about visualizing, and how you can use visualization to understand the distance between the present and the desired future. This newsletter is about playing chess, and a different kind of vision - chess vision.

Chess vision means you can see what will happen. You can see moves ahead. You can imagine not only a sequence of moves, but also how the board looks when those moves are made, so you know the outcome and can play for the best outcome. To be an adequate chess player, you need to see four moves ahead. Rated chess players see far deeper.

There are leaders in this industry making decisions without looking moves ahead. Here is an example: Microsoft and Intel think of a great idea (actually, Steve Jobs though of this great idea around 1982, but hey) called the Simply Interactive PC. It is a sealed box. It has two external busses, one fast, and one very very fast, that eliminate the need for almost all internal expansion capability in desktop PCs. It is an information appliance that can be packaged attractively for people like my mother, who is currently better off with her old Macintosh. In fact, the Simply Interactive PC is a Macintosh - the original Macintosh, before Apple lost its way and made the keyboard too big, the boxes ugly sheet-metal contraptions, and the OS over-complex (but that is a much longer and more woeful tale of lack of vision).

The Simply Interactive PC would revolutionize the peripherals and enhancements business: instead of having to open your PC, you would put the simple, medium-performance devices like modems on one bus - the USB bus - and fast peripherals like DVD drives and digital camcorders on the other - the 1394 or FireWire bus. And you can "hot" plug and unplug to your heart's desire. This would greatly expand the market for enhancements and enable innovative new uses for PCs, along with the hardware and software enhancements that go along with these new uses. Home video and music editing and other creative applications were high on the list of possible new high-growth areas. More important, another tier of customers would be added to the market, those that find current PCs a noisy intrusion of a complex office machine into their homes.

What happened to the Simple Interactive PC? It turned into the flop du jour: the NetPC. With a crude false attack, Sun lured the PC world off its plan to fundamentally expand the PC market. Sun's JavaStation and the even more unicorn-like NC from Oracle (a hardware maker?) stampeded the PC industry into defending its business desktops from this unlikely incursion. The Simply Interactive PC, the PC for my mother, became the defender of the business desktop.

That's what comes of thinking one move ahead: Sun and Oracle posit a device, and the PC industry moves, with a heavy hand, to squash it. Trouble is, like chess pieces, technology strategies unfold in several moves. Microsoft and the PC makers, seeing an attack, failed to see the intermediate moves, failed to see the weak ending position the attack would have, and in the process incurred an opportunity cost: all the millions of people who would have bought a Simply Interactive PC, and all the nifty toys to attach to it, have none to buy. Apple was left breathing room. My mom is left with her old Mac. I don't have a good excuse to buy a DVT camcorder ("Honey, we'll want to edit down those tapes of the new baby for great-grandma."). And we're all waiting for smart set-top boxes when the network interface and decoder hardware for all that fancy digital cable should all just be attached to the 1394 bus of one of these Simply Interactive PCs, and that the set top box should not be smart at all.

Meanwhile, on the other side of the chessboard, Microsoft is trying to stuff the Java genie back into the bottle. Unlikely. Every muffed opportunity has a cost that spans the chessboard: What if Simply Interactive PCs were out in the market now, connecting to set-top boxes, providing Internet connectivity in homes, and, probably, spilling into the parts of the business market where they were really needed? How much better positioned would Microsoft be to promote ActiveX as the power behind making the Internet more interactive?

Chess, and business competition, is all about delayed gratification. If you have no patience, if you are prone to first-order analyses, you'll never see even four moves ahead. But even if you do have the discipline to see four moves ahead, will the big companies who can hire firms like Bain and McKinsey to do the looking ahead for them still beat you? Don't worry. Big companies are run by gangs of vice presidents waiting for other vice presidents to make a mistake and back-stab them. It's hard to look ahead and watch your back at the same time.

And now, it's time to play computer telephony chess.

Move one: Create the UnPBX. Put a switch fabric on an add-in card and you have a telephone switch inside a PC server. Sell it to a customer.

Move two: two years from the time the customer installs the server, the server is looking long in the tooth. In three years, the server needs replacing. The new server has only one ISA slot, the rest are PCI, and that UnPBX is implemented on four (?!) ISA cards. Ow! Later that day, while grousing to a friend over a beer, customer learns his friend has a 15 year old PBX, and no plans to replace it.

This is a bind I've mentioned to a lot of my colleagues, and the best answer I've gotten so far is "Yeah, well, uh, people will start replacing their PBXs the way they replace their servers." Right. PBXs are more like plumbing than computers, despite the superficial resemblance. Imagine your plumbing incorporated a PC.

Look at this, instead:

Move one: Customer with conventional PBX, and a network of PCs, wants to use Internet telephony.

Move two: vendor of PBX enables customer to remove the control processor of the PBX and replace it with a link to an external server over - pick one, depending on your preference and the system size - Ethernet, ATM, 1394, or USB. Now the external server runs the PBX, the control software of the PBX having been ported from whatever grotty old OS the PBX maker had been using to NT. The server also has access to the PBX's internal bus, and can "export" and "import" a large number of slots in the switch fabric (total depending on link technology, but could be as high as the total bus bandwidth of the PBX) to the server from the network and vica versa. So the server is also a gateway from the phones on the customer's desktops to the Internet.

Move three: Three years later, customer upgrades server. Now the server can do more: gatewaying voice calls from telephones into online collaborative conferences, simulating a conference bridge, fax in software, etc. Customer cuts over to new server with a less than one-minute outage - the time it takes to unplug and plug in the link cable.

Move four: Two years later, customer wants still more: adds a DSP resource inside the server for more faxing, more conference bridging, etc. Does he need a separate bus inside the PC? No, this new server has 64 bit fast PCI, and dozens of voice channels are a drop in the bandwidth bucket. Later that day, over a beer, customer makes his friend with the 15 year old PBX insanely envious of his operating efficiency.

Am I Altigen bashing here? No. I want an Altigen. It's cool. I like it. I'd use one knowing full well that I'd toss it in three years. But Altigen themselves have to consider what happens to the bulk of their customers without my sense of adventure (and I still wouldn't put a PC in my plumbing).

PBX makers have to realize that their product formulation depends on them having the right chess vision. Misadventures like in-the-skin PCs, PBXs in PC chassis, etc. are wild guesses, seeing no moves ahead. Seeing the product lifetime differential between PC servers and PBXs is the first step to seeing far enough ahead to deliver the right product. What if the Lucent Partner small business switch came with a USB interface, a built-in voice processing resource, and a Unified Messaging software package you could install on any PC that you would like to attach to this system? Such a configuration would put that hoary old epitome of conventional small business switches right on top of real customer needs.

Some other examples of not seeing moves ahead:

Move one: Apple offs their clone vendors. They might be right that they are subsidizing those vendors, but the numbers are open to other interpretations as well.

Move two: Loyal customers, fanatically loyal customers, get really pissed-off. Did anyone calculate the value of all those Mac fanatics - people who had been actively urging people to buy and not pirate the most recent software update? You cannot buy brand loyalty like that. You can only spend it. Wasted.

Move three: Motorola, Apple's CPU vendor says: "We're probably out of the desktop CPU business." Holy supplier risk! Apple better be very sure about the timing of the transition of their customers to Next technology. They may face a very bleak period where they bleed market share due to un-competitive CPUs.

Move four: Apple launches Rhapsody, and licenses it to… Yeah, right. Want to see a venture capitalist die laughing? Hand him a business plan for a Rhapsody-based desktop computer product.

What to make of this? Steve Jobs is very intelligent. So it is possible that he has a very obscure plan to win. Trouble is, plans for winning are usually not obscure. They are usually simple, like Wal-Mart using computers to keep useful things in stock. Or Ford using TQM to reduce defects in cars. These are very simple strategies. The chess vision involved in executing these strategies is in the fact they take multiple moves to implement: Wal-Mart had to get their suppliers to change the way they do business, and not all were ready to do so at once. Ford had to take it on faith that Americans, who, the numbers plainly showed, bought lots of crappy American cars in the 70's, would respond better to quality. What Apple should have done, and may not have been able to, under any circumstances do, is face the fact that they would have to ride out their current licensing regime until they could replace Mac OS with Rhapsody.

What is the value of honing your chess vision? It gives you the ability to put in place and defend a strategy that takes multiple steps to implement. Remember that the strategy itself is usually very simple. The multi-step execution is where people get lost. Especially when put under pressure: "People love vinyl tops, 'opera windows' and fake wire wheels. Look at all those cars we sell!"

Look ahead and prosper.

Copyright 1997 Zigurd Mednieks. May be reproduced and redistributed with attribution and this notice intact.