All Posts in Strategy and Tactics

|In: Industry Commentary|Last Updated:

I’m jealous of political pundits. They get to write about global warming, economic protectionism, and cloning (for example) without any particular expertise in the relevant disciplines, influencing public policy in the process.

Not me. All I get to write about is how to run IS. My writing about Zippergate, for instance, would be about as classy as the Hollywood actors who make political speeches during the Academy Awards (although that may be preferable to the current practice of thanking everyone in the telephone directory).

That’s why I’m so thankful for Department of Justice vs. Microsoft. It lets me comment on a major issue of public policy without straying from my purported area of expertise. Here’s my comment: As of this writing, Microsoft seems to be counting on one of three legal possibilities: Either 1) It isn’t really a monopoly because even though it is right now it won’t be forever so that makes everything OK; 2) the antitrust laws are a bad idea so don’t enforce them; or 3) even if it’s found guilty, the penalty won’t hurt very much. The “consumers haven’t been hurt” argument is, of course, irrelevant grandstanding – the question is one of using its monopoly for competitive advantage, not one of price-gouging.

The biggest impact of this trial, of course, is all the free publicity it’s given Linux. Ain’t irony grand?

Last week’s column presented a simple formula for predicting the success and failure of new technologies and technology products, using past products as examples. This week we’ll apply it to the current technology scene, starting with Linux.

The formula, you’ll recall, was customers/affordability/disruption – a successful new product must do something worthwhile for the customers (the people making the buying decision, as opposed to the consumers, who use the product); it must be affordable; and it must not disrupt the current environment. (Disruption, by the way, is a big reason companies like Microsoft, which dictate the architecture in particular environments, have the incumbent’s huge advantage.) Let’s start predicting!

Linux (as a server) – Main customers: Webmasters and underfunded system administrators. Benefit: Runs reliably and fast. Affordability: Free, or nearly so. Disruption: As a Web server or file-and-print server, it integrates invisibly into the network (unless Microsoft can make it disruptive through proprietary server-side innovations like Active Server Pages). Score: Perfect – Linux is a winner.

Linux (on the desktop) – Main customers: End-users. Benefit: Fast, less-crash-prone PC. Affordability: Free except for the (not overwhelming but not trivial) time needed to learn it. Disruption: The office suites for Linux don’t reliably read and write the Microsoft Office file formats – that is, there’s a significant delay before the Linux suites catch up to each new Office release, and even then they’re glitchy. Score: Iffy.

Personal Digital Assistants (PDAs) – Main customers: End-users. Benefits: Lightweight, carries around essential information, IS doesn’t get to say what users can and can’t do with it. Affordability: Very. Disruption: Unless an installation goes south, they’re invisible to IS. Score: Perfect – PDAs are a winner.

XML – Main customers: IS developers. Benefits: Universal, awesomely adaptable file format. Affordability: Open standard, learnable without needing tensor calculus. Disruption: A complicated question. As meta data (think data dictionary on steroids) there’s no significant installed base to disrupt. As an office-suite file format, either Microsoft’s XML tags will do the job for everyone or it won’t get off the ground. As a replacement for HTML it’s highly disruptive until it’s built into the browser. Then it’s nondisruptive. Score: Very High – XML has enough different applications that it’s bound to succeed in at least some of them.

Java – Main customers: IS developers. Benefits: Nicely designed object-oriented language, automatic garbage collection, possibly portable (the jury’s out on this one). Affordability: As affordable as any other programming language. Disruption: A complicated issue. For established developers it’s disruptive – having some of a product in Java and the rest in a compiled language is messy, and probably won’t happen. For new developers it’s non-disruptive, except for the performance issues compared to any compiled language. Score: Adequate – Java will become just another programming language.

Well, there you have it. Like a good math text, you now have a formula and examples. Go forth and prognosticate.

Comments

|In: Industry Commentary|Last Updated:

Among the many careers I want to try before I die — televangelist, rock star, college bar owner, psychic — is founding an organization to help a neglected group I’m part of. I’ve been planning it for some time: Sarcastics Anonymous. I’ll start the meetings.

“Hello, my name is Bob and I’m sarcastic,” I can hear myself say.

“What else is new?” I hear the assembled group respond.

I’ll probably never get SA off the ground. I need a 12-step program, but I’m stuck at two: (1) Notice something; and (2) snicker at it.

I need to stop before I sneer again. I’m especially vulnerable to self-improvement books. I want to write The Three Habits of Somewhat Effective People, and a satire of Life’s Little Instruction Book written in the mangled English you find in the instruction manuals of some Japanese products.

Most of all, I want to write When Good Things Happen to Bad People and its sequel, When Good Marketshare Happens to Bad Products.

Ever wonder why bad products win? The quadrant charts beloved of expensive market research companies don’t explain it. Neither does product quality, except to break a tie, nor customer stupidity.

Three factors determine the success or failure of any new technology product: What it will do for its customers; its affordability or lack of it; and how much disruption it causes. Here’s how it works:

Customers: Customers make buying decisions, as opposed to consumers, who make use of products. The two are distinct. For a new product to succeed it must do something useful for the customer … in other words, the customer and consumer must be the same.

Think about the ill-fated OS/2 in this light. IBM defined the IS director as its customer, positioning OS/2 as a stable platform for client/server applications. The customer wasn’t the consumer. That was the end-user, for whom OS/2 provided negligible benefit.

Affordability: New products are risks. High prices make them riskier. Consider Lotus Notes, so expensive in its early, critical years that “Doing Nothing,” its only competitor, had about 50 times more market share.

Disruption: Disruptive products, those that don’t get along well with the installed base, are guaranteed losers. Bet against them.

There are two ways new products can be non-disruptive. One is to avoid touching the installed base. Early PCs were like this. Alone they sat on end-user desks (end-users were both customer and consumer for these affordable devices), connected to nothing.

The other way to be non-disruptive is to integrate smoothly into the existing environment, creating a smooth migration path. The early NetWare, for example, was invisible. F: behaved much as the C: end-users already knew; LPT1: simply went to a better printer.

Windows was another spectacularly successful example. Version 2.0 looked like a GUI toolkit to developers, saving them work. End-users didn’t even have to install it — it shipped as a run-time environment with Windows applications, loading and unloading with PageMaker, Word, and Excel. Version 3 remained non-disruptive but in a different way, running DOS applications transparently instead of only loading when needed.

Oh yeah, those versions logged onto NetWare just fine — mostly, they didn’t know it was there.

Compare Windows with the early OS/2, vying for control of the GUI operating system market space. OS/2 made you throw out your hardware. IBM told you so, hinting broadly that OS/2’s true power couldn’t be unleashed without Micro Channel. Your software? OS/2’s DOS “compatibility box” was useless, so you had to start over with a new OS/2-specific word processor and spreadsheet. Log onto NetWare? Sure, using a new, immature technology, not the tried-and-true drivers everyone already knew.

(IBM did a great job, though, compared with Next. Next had no identifiable customer or consumer, cost a ton, and integrated with nothing at all. It had “Market Failure Guaranteed” stamped on the side of every box.)

“What makes you the expert?” I hear some of you ask sarcastically.

Nothing. I’m applying a old, well-worn marketing principle: Make buying from you easy.

So don’t sneer at me — save it for the venture capitalists who squandered billions ignoring this simple rule.

A fool and his money, after all …

Comments