Many Linux enthusiasts believe that Linux has grown through technological superiority, and the superior morality of open source software. Unquestionably that’s part of it. Yet it is not the end of the story. There are a thousand ways a new idea can fail, few to succeed. The path to market can be narrow and winding.
It can also be surprising. More than one Linux user has pulled their head out of an Alpha one fine morning, picked up the local newspaper, and found the penguin sitting smack dab at the top of the Technology section. Why now, they wonder; we’ve been around awhile. It’s a question worth asking, since opportunity doesn’t hang around forever.
From a background of marketing, media, and long-forgotten code, these are a few thoughts on where Linux stands with the press and in the marketplace and where it is bound.
(I use "Linux" in the popular sense, to describe the kernel, libraries, and GUIs packaged in distributions.)
You don’t give deep thought to buying a bag of peanuts. If the price is right, and they look OK, you buy.
That changes when large sums of money change hands. People with broad experience in businessand realityknow that customers willingly give their money to those they perceive as winners. Conversely, they do not give money to losers; it runs against the grain. This attitude cuts across all strata of society, and colors not only commerce but daily decisions in all walks of life.
Mainstream wisdom would challenge this with theories like "Americans love an underdog." There is little evidence for this in a reading of American history. It is a history of extreme ambitions, riddled with a mania for acquisition. Through the 1800s that mania was mainly focused on land. By the 1900s the thrust had almost entirely changed to money and market dominance. It is this arena that Linux is entering, with both media and customers measuring the winners by cash and glory.
The mania could be seen at its worst in the now-defunct Newsweek column "Winners and Losers." Newsweek, of course, is merely reflective fluff. It takes a Kurt Vonnegut to deliver a more brutal and accurate assessment. In a 1968 article in Ramparts magazine, on the Republican presidential convention, he wrote "... The Democrats keep saying, ’We are the people.’ But the people know they are losers, and losers don’t vote for losers. They vote for winners." (Thirty years may have corrupted a couple of neurons, so correct me if the quote is off.)
Why then, would the media embrace the penguin? Because Americans are fantastically quick to embrace an underdog with a shot at the championship. In the movie "Rocky I," the crowd cheers instantly when Rocky drops the champ. An American audience has its nose to the wind like no other.
They are equally quick to turn away from a loser. The real battle is still ahead.
Companies like Red Hat, VA Research, and Linuxcare are now bedeviled or blessed by waves of media types identifying themselves as "tech reporters." As a rule, they’re not technical at all. While not everyone was born to write code, or coded to write, it would be nice to think tech reporters can insert an IF function in a macro. That’s not realistic. In a fairly common example of media practices, a Wall St. Journal reporter on the search engine beat recently acknowledged ignorance of www.searchenginewatch.com, probably the world's leading informative site about the search engines. Real tech reporters hail from LinuxWorld, from Smart Reseller, or even PC Magazine.
A reporter calling from a newspaper or mainstream magazine is far more likely a business reporter on the hi-tech beat. Out there in medialand, there is someone on the United Airlines beat, the Boeing beat, and the Hasbro toys beat. There is probably a pork bellies beat. (Major newspapers will actually have a full-time reporter on the Yahoo! beat or AOL beat. Linux is not currently a beat of its own, though sometimes it feels that way.)
Sexiest of all is the high-tech beat. Within that field there are subcategories of stories ("articles" to those who don’t watch TV).
Sexiest of the sexiest is mergers & IPOs. All popular media outlets cover these, though smaller outlets will tend to use stories pulled off the newswire. (Which is why you see so many stories from the San Jose Mercury News. They move through the KRTN newswire, and pop up in papers across North America.)
Shows, conventions, and events also generate coverage. Stories about artificial events such as the recent Refund Day are somewhat puzzling to people who read a newspaper for information, but events are part of the media's formula. In fact editors were mildly awed that 100 people led by Obiwan Kenobi were able to go national on the newswires.
For this reason the substantial sums blown on the LinuxWorld convention make some sense. An event gives the media something to hang their hat on; in media lingo, a "newspeg." I say "blown," because few conventions pay out in hard, trackable sales revenues. The theory for setting up a booth usually comes down to "cause ya gotta." (In this case it makes sensepresupposing that the only viable path for Linux to become the standard OS is to follow the dot-com IPO route. But that may an Ameri-centric view. MS never locked up users’ minds quite as well in the LDCs.)
Another variant is the high-tech equipment story. These will be the reviews of Linux boxes the ones that come with the OS and GUI pre-installed. When there is a significant choice of competing products, the newspapers will do "roundups," comparing the products. The reporters who do these have home offices stuffed with scanners, VR technology, Palm Pilots, cellular modems, and the latest printers. They still don’t write code, but they are willing to get out the screwdriver to add an external peripheral.
Another press venue is the Question & Answer format. These columnists vary. Some clearly cull their information from mainstream PC magazines or technical friends, and then regurgitate it to the readers. Others are fairly technical, and could install a Linux distribution without trouble. Some of this group are PC/Mac magazine writers, migrating into mainstream media.
The long and short of this is that different reporters after different stories want different information. If there is a pattern in Linux evangelism, it is for the hackers to talk tech, and the future suits to talk mondo stupendo. The writers, however, vary according to their beat, their technical knowledge, and their story assignment. When the information doesn’t fit in the story, they move on to another source.
(In most cases, the stories will appear oddly irrelevant to a Linux user much as a USA Today story about how much a movie grossed on the first weekend will leave a reader wondering where the review is, anyway. Nevertheless, they’re good publicity.)
Some of the more effective publicists such as Linus Torvalds or Eric Raymond can obviously switch-hit between technical and business stories. Most Linux users don’t have the experience or open-mindedness to do this. When they also don’t have the presence of mind to redirect the press to the right person, coverage will be lost.
The tendency to subcontract publicity work is natural enough. Companies entering the arena usually have no clue how the media works, and PR work seems ideal for outsourcing, since it’s cleanly separated from tech, administration, and marketing. Or is there a clean separation?
Linux companies using professional PR firms inevitably slide towards the business/financial/hype angles, since the flacks can’t understand the technology. Overall this is OK, since that is what most of the stories are about. However, the mindset will kill numerous technical articles, as frustrated Q&A columnists and (real) technical writers, seeing deadline approach, wander off to more easily written stories.
Blocking media access can create a bigger problem when it kills mainstream business stories. By screening the press away from the quote they need, a subcontract PR agency will almost inevitably kill stories that just needed a quick quote or two from someone inside the company. (Quotes from PR agencies are rarely usable; editors frown on them, and most reporters learn to avoid them.)
The strength of a PR firm is that it demonstrates a company is a real player, not just a wannabe. Of course high-quality in-house PR can achieve the same thing. Amazon wisely hired on a Hollywood veteran; she is predictably shameless, but also experienced, and along with her gift of gab is able to quickly refer reporters to the right person.
Generally PR firms call to mind Machiavelli's advice that ". . . if anyone supports his state by the arms of mercenaries, he will never stand firm or sure, as they are disunited, ambitious, without discipline, faithless, bold amongst friends, cowardly amongst enemies . . ." (The Prince, Chapter 12, Luigi Ricci translation)
The short-term solution is to hire a PR firm for the proactive workactually chasing the pressand make someone inhouse the traffic controller for incoming reporters. Since the incoming reporters already have a story in mind, this requires no publicity skills; just the brains and company knowledge to route the reporters to the right source promptly.
In choosing a penguin, it would seem Linus Torvalds picked a media winner. (This is not jumping to the conclusion that it’s a market winner; media success is not automatic market success.)
In art form it will obviously fly within the Linux community, and doubtless minor fortunes will emerge from the fuzzy penguin toy market.
The press should also pick up on the penguin. Initially there will be two attitudes. One will be that the penguin is ever so cute, let’s put it on top of the article. The other will say that the penguin is just too cute, let’s not. If Linux keeps capturing market share, there will be drawings of massed penguins marching out of the South Pole, presumably headed for Redmond.
Eventually it will taper off, but by then every copyeditor will have a small penguin logo in their file cabinet, to tack onto short capsules (these are the business snippets seen along the left side of a newspaper’s business section, with the company logos next to many of them).
Print will also sober up eventually. At first there will be a spate of writers having fun with animal imagery: "The march of the penguin," "Can penguins fly?," etc. If the penguin can land a few more good blows on the competition, coverage will move towards the business/sports news style, "Well, Bob, I didn’t hold out much hope for the challenger, but that left hook in the fourth round sure shook the champ. Do you think he could become a serious contender?" This is sometimes known as a little train that could story.
"Open source" works in part because it’s incomprehensible (reporters aren’t going to understand compiling without lengthy explanation). The words "open" and "source" are also good in themselves. Open speaks of freedom, of the open road. Source speaks of wisdom, like a wind from Delphi.
"Free," on the other hand, is vile, vulgar, and low class. That it’s the root of "freedom" is irrelevant. That it’s an incredibly powerful word in direct marketing is irrelevant. The press will not like it, and it won’t play with the suits.
That Linux is actually the kernel and GNU the OS is not likely to make the newspapers, except as an in-depth story in the likes of the San Jose Mercury-News or the N.Y. Times. Up-market publications like the Financial Times or The Economist can also handle the concept. It is difficult to see the distinction making it into a story about the LinuxWorld convention. Witness the media’s inability to distinguish between DOS and Windows, or hackers and crackers; a distinction like "Linux kernel with a GNU OS and Gnome GUI" is well beyond the popular press. It is likely that without a clear and pronouncable name, Linux would never have made it this far.
If GNU and Gnome and KDE are to become recognized beyond the open source community, they probably need new names. Outrageous but true. Emacs is a decent name, though editors would see it as being confused with Macintosh.
New to marketing, many hackers have grabbed at the handiest marketing concept to guide their plans. In the late 1990’s that happens to be "branding." It is difficult to say when this became a verb. The current incarnation seems to encompass several quite different theories, the three most obvious being brand awareness, brand loyalty, and inertia marketing.
Brand awareness is what suits (and non-suits) buy to please the CEO. Formerly known as "getting your name out there," or "name recognition," it means spending large sums on ads that eventually cause people to say, "Oh, yeah, I’ve heard of them." Your advertising agency thanks you for your continued support.
Brand loyalty, a more elusive but useful quality, occurs when customers have a set of beliefs about a product, and will continue to buy it in the face of new competition.
Microsoftperhaps surprisingly to a Linux userhas substantial brand loyalty. In a world where there are more followers than leaders, a huge number of computer users have a slavish adulation of Microsoft, simply because Bill Gates is listed as the world’s richest man. Nor does the "bully" reputation bother everyone. Social Darwinism is alive and well among conservatives and authoritarian personalities, and there are plenty of these people.
Inertia marketing is more powerful than either brand awareness or loyalty. Bill Gates probably never expected to have a huge range of humanity adulating him. So Microsoft made themselves ubiquitous, indispensable, pervasive. And when someone’s gas gauge starts pegging empty, you have need piled on top of inertia. Inertia, need ... and then branding: the combination that made MS what it is or isn’t today.
There is a tendency among those new to any field to emulate success stories. It’s a reasonable attitude, but in marketing a dangerous one.
The problem is that few people actually know how their role models succeeded. In the computer industry, at least the founders and observers are around. When companies start likening their plans to Henry Ford or Andrew Carnegie, the comparisons become unlikely. The witnesses are gone, and the times have changed. Yes, ketchup captured .147% of the U.S. grocery dollar, last time I checked. But Linux ain’t ketchup.
"Make a great product, and give great service!" is not a marketing plan. In fact
it has no meaning, without knowing the product; timing; place; price; channels; buyer psychology,
needs, wants; competition; overhead; cost of capital; regulation; promotion costs; and a
few dozen other things. Without pummeling any tech company in particular, there are a lot of
simplistic theories in action. Good plans are usually simple at core, but not in the details.
It is no secret that hardware manufacturers have been lining up for the Linux bandwagon. Why not? Fifteen years of Redmond is enough to make a convert out of most insiders, software or hardware.
The courage of their conviction remains in question. Many of these companies are run by suits. The founders are long gone, and the suits who now hold the thrones may be treacherous, but not necessarily courageous.
Companies still run by their founders can be quite different. Some of the founders are certainly treacherous. On the other hand, most founders are also bullheaded. Old injuries are not easily forgotten by ambitious people, and one can be sure a significant percentage of founder’run companies are sharpening their knives for "the beast of Redmond."
Not all the knives will be pointed at Redmond. Intel has also made enemies, which raises the question of the chip market. The Federal assault on Wintel, the increased Microsoft bashing in the press, Intel’s serial number feature on the Pentium IIIand to a lesser extent the rise of open source codeopens a door for AMD and Cyrix to position themselves as something other than "cheaper." While Intel is now climbing on the bandwagon, AMD and Cyrix have stronger motivations: they could gain market share. They may see Linux as a significant vehicle to that share.
Bruce Perens of the Open Software Initiative has expressed concerns about a legal assault on open source patents. I’m not current on software patent law, but it may be realistic to think that Microsoft would unleash a massive assault on individual patents and copyrights, perhaps filing endless low-profile lawsuits in different courts.
Yet they probably won’t as long as the Federal anti-trust suit is open. This broad suit is not cast in stone. You can be sure Federal lawyers are constantly looking for new evidence to introduce. This is not a personal guarantee, but an educated guess based on knowing countless government lawyers. When the lawsuit ends, all bets are off.
(Perens has also expressed worry about IBM and Lucent making patent assaults. This is another concern, considering the two companies’ broad patent bases.)
January, 1999 - newspaper editor says, "We’ll probably do something else on Linux in 7-8 months." Within the month the editor assigns 2 major stories on Linux.
February, 1999 - businesswoman says, "Linux! I want to buy stock! Where do I find the company?"
March, 1999 - magazine editor asks for a story on "the Linux company."
Linux can win in the marketplace, both server and desktop. But the penguin may have to win both battles, because capturing the minds of the suits may require both victories. It is possible that without both victories, Microsoft can still drive NT to eventual network dominance.
It’s easy to focus on the server war. It can be won with the product, the backing of IT people, and the press. And if IT people are not always true believers, at least some of them can print "Hello World." So at worst they are kinda OK, and comfortable to work with.
The problem is, the press may not stick around if Linux isn’t winning the desktop war.
To win the desktop war Linux needs:
In large part success depends on whether commercial Linux companies alienate the open source community. It would be easy to mock the programmers who toiled to make Linux work. It would also be reckless, as they remain the backbone of the open source movement. The suits grabbed at encryption too early, and lost the backing of the developers. Encryption will still become a giant industry, because it’s needed, because big business will pay enough to maintain R&D, and because MS isn’t after it. It’s not a parallel for Linux, but grabbing too early is a piece of history worth noting. So far the future Linux suits have been careful, indeed pious. A few IPOs followed by arrogance may change that.
It also depends on whether a reasonably broad reach of Linux users will accept the public. Most are gracious with newbies. Others are less so. Unfortunately, "They’re stupid" is not a viable sales approach. Nor is it an accurate one. Some smart people can’t use computers very well.
It was discovered at SRI and PARC that it’s a hell of a lot harder to design something simple. Likewise Wozniak never looked down on usability, or even good user manuals. If Linux is to go mainstream, that mindset may be the price of admission.
January 11, 2001 The beat goes on ... people keep visiting this paper, so here’s my two bits:
The popular media has all but forgotten Linux, so even without the 2000 NASDAQ drop, the time for Linux IPOs is over for awhile. It’s bottom line time again. Linux companies that were founded on a grow-fast go-public business model are now hurting as they try to morph to a new business model, if they can figure it out in time. Since there are viable business models for Linux, the current game in Linux consultancy is "last man standing." Like boxing, the victors may be bloody, but they get to take the gold belt home.
Meanwhile Linux is spreading into industry. It’s beginning to dawn on some of the companies that there is more money in customization than in tech support.