Wednesday, August 10, 2005

August 2005 Early Indications I: Remembering Windows 95

It's a slow time in the technology industry. Breakthrough innovations
are few and far between: the iPod is almost four years old, and it's
hard to point to anything very interesting since then. Because
revenue growth has slowed, mergers and acquisitions have become the
main order of business at such companies as Oracle and, for a time,
HP. Venture capital is increasingly migrating to biotech, physical
security, and other sectors only tangentially related to computing.
The industry could use an injection of energy, activity, and not least
important, revenue.

Given this state of things, everyone is watching Microsoft, which is
preparing to launch a new operating system next year. Last month
merely changing the name from code (Longhorn) to product (Vista)
devoured a lot of attention, and more recently a stripped-down version
of the product shipped to beta testers. The product has been a long
time in coming, and the scope has been managed downward in several
respects. Nevertheless, both Microsoft and the industry more
generally see Vista as a potential jump-start very much in the same
category as Windows 95 ten years ago. Because Vista represents the
first opportunity in over ten years to begin with a "clean sheet of
paper," unlike Windows 3.1, 98, ME, and 2000/XP, Bill Gates has
repeatedly linked the two products in public.

Before looking at whether that association is warranted, it's worth
remembering just what Windows 95 brought to market. In 1994, loading
a browser onto Windows could be complicated by the operating system's
lack of Internet Protocol support. DOS prompts were very much a
day-to-day reality. File names were limited to eight letters, and
CD-ROM support was spotty. E-mail was used only by fringe populations
rather than being nearly universal. Adding hardware was more
difficult than it needed to be, multitasking was nearly impossible for
both processing and user interface reasons, and multimedia computing
was, again, the province of only a small subset of users.

Windows 95 changed all of that. Even before Gates' famous "Pearl
Harbor" speech helped turn Microsoft into an Internet-aware company,
Windows 95 made Internet connection, through both browser and e-mail,
a mass phenomenon. Multimedia, too, became an everyday event with
better hardware support (including CD-ROM drivers). Overall
usability, despite the initial confusion at using a "Start" button to
shut down a machine, was enhanced by deeper camouflaging of the
command-line layer, longer file names, and plug-and-play peripheral
support. Finally, the operating system kept pace with Intel's chip
performance and supported more realistic instances of multitasking.

The public responded. In the quarters immediately following the
launch, retail sales of Windows 95 software soared, augmenting a
strong increase in OEM sales of pre-loaded operating systems.
Responding positively to improved networking support and promises of
enhanced manageability, corporate IT organizations spent at record
levels: Microsoft's operating systems revenues jumped from $1.5
billion in fiscal 1994 to $4.1 billion only two years later.

What might we deduce about the prospects for Vista based on the
Windows 95 experience? First, it's hard to see a parallel burst of
initial interest, with or without a Rolling Stones commercial.
According to Microsoft, the benefits of Vista fall under five general

-Deployment (for organizations managing large rollouts)
-Performance (including better power management and faster boot up)
-Management of distributed users' machines

These categories of improvements are clearly aimed at corporate buyers
more than individuals. Most of the things a consumer-grade user will
see - including better desktop graphics, and RSS support within
Internet Explorer - already come standard in Mac OS X. Backward
compatibility will be substantial, to the point that many Vista
improvements (including the IE browser) will be available as retrofits
to Windows XP. These upgrades will also slow Vista adoption.

Here's another way of thinking about the comparison. In 1995,
Microsoft turned the telephone network into an extension of the
computer, or vice versa: between them AOL and Windows 95 made the
Internet a household utility. In 2006, no parallel leap into an
adjoining domain - think of home entertainment, specifically the
television - will be supported. Bill Gates longstanding prediction
about widespread adoption of a voice and speech interface to the PC
will be addressed with Vista support, but even given a powerful
standard processor configuration at its disposal, Vista still won't
make masses of people retire their keyboards.

In short, Windows Vista looks like a solid product for corporate
purchasers, but the lack of "gee-whiz" and "I've always wanted to be
able to do that" desirability will prevent end-user excitement from
reappearing the way it did ten years ago. An industry in search of
the next big thing will probably have to keep looking.

Monday, August 01, 2005

July 2005 Early Indications II: Virtual Trust

[Business notice: I have officially formed a company, Still River Research, to deliver consulting and analysis services. The website ( is now out of beta after generous suggestions from several newsletter readers. I am now booking projects for the fall; please notify me if I can be of service.]

Security stories currently dominate much of the news. Between London, the Patriot Act, data thefts and losses, and renewed efforts to mandate identity cards for immigrants, it's difficult to help but feel that the world is a scary, dangerous place. My focus here, however, is on a near neighbor to security: trust, and how it can be both reinforced and undermined in new ways via digital networks.

It doesn't take long to see how various online efforts attempt to prove their trustworthiness:

-eBay relies on collated word of mouth to label bad apples and reassure good citizens. The company's institutionalized reputational currency ("view my 100% positive feedback!") is not patentable yet constitutes an enormous barrier to competitive entry.

-Some social network and dating sites use acquaintances as proxies: "you don't know me, but you know Mike, and Mike knows me, so I'm probably OK." As the Spokes and Friendsters of the world have discovered, trying to scale friend-of-a-friend trust is neither obvious nor cheap. When was the last time you used one of these services and could honestly say it was overwhelmingly positive?

-Other dating sites rely on the objective authority of social science. At eHarmony, potential daters are greeted by "relationship expert Dr. Neil Clark Warren" who has built a "detailed questionnaire measures the intricate facets of a person, including the 29 dimensions that are most important in relationship success." Not only that, an American Psychological Association conference included a paper that suggests that eHarmony marriages are happier than marriages built on other matchmaking techniques.

-Some entities have had a difficult time recreating the trust they built offline in new media. According to Lawrence Baxter, chief e-commerce officer at Wachovia quoted in the July 21 Boston Globe, the bank can no longer use e-mail to communicate with customers because phishing attacks so skillfully recreated the look and feel of official correspondence that customers routinely delete real messages. Cost structures used in the online bank's business case, meanwhile, almost certainly are rendered obsolete by the need to revert to physical mail.

-Amidst all of the 10-year celebrations of e-commerce sites (eBay, Amazon, CNet), some longtime readers may recall our discussion of Encyclopedia Britannica, which was nearly wiped off the map after over 225 years of operation. The company still exists, still publishes multi-volume hard-copy products, and recently announced it had re-formed and upgraded its panel of experts. That body, once home primarily to white males, now includes four Nobel laureates, two Pulitzer Prize winners, and a much more representative cultural makeup. Significantly, the last meeting of the board was ten years ago.

Several conclusions emerge:

1) Trust pays: eHarmony says they get 10,000-15,000 new members a day, each of whom has spent between $50 and $250.

2) Trust is expensive to build. As I searched for a new cell phone, Staples referred me to an outside vendor, but the vendor's site retains a Staples logo at the top, with the reminder that Staples will stand behind any transactions. The vendor's own site, with no such guarantee, sells the exact same service plan and phone for $50 less. Given the failure rate and overall dissatisfaction with U.S. wireless carriers, that $50 insurance looks very appealing.

3) There's a fallacy that identification can routinize trust: TSA screenings assume that someone with a driver's license that matches her face won't try to do anything bad to the aircraft. Conversely, someone who doesn't provide ID is kept off the plane: former Sun Microsystems employee John Gilmore is in federal court challenging the unwritten and/or secret law (nobody has yet produced it) that states that an "internal passport," as he calls it, is a condition for public transportation. (Here's the Gilmore site.)

4) The Britannica case, in its contrast with Wikipedia, highlights a particular dynamic on the Net, that of open vs. closed credibility, or trust if you will. Much as "many eyes make bugs shallow," as Eric Raymond argued in The Cathedral and the Bazaar in reference to open-source software, Wikipedia establishes trust in the volume of researcher-reader-editors who will spot and fix errors. Unlike the Staples model, money is less effective than reputational currency - the same stock of "funds" that makes eBay work.

Britannica, on the other hand, seeks the credibility of the few: the Encyclopedia's editor stated that "At a time when vast quantities of questionable information are available on the Internet and elsewhere, rigorous and reliable reference works are more important than ever." They are, but Britannica has a lot to answer for: the BBC reported that a 12-year old boy in London found five errors in two entries. Add to the errors the cost to fix paper editions, and the lag between error detection and correction - how many readers will propagate errors in the interval?

5) In the physical world, institutions can convey cues that reassure patrons of their solidity and presumably good intentions: marble pillars on a bank, brightly lit colorful plastic in a strip mall, even flight attendants' and pilots' uniforms. Online, Wells Fargo, Target, or Delta can't convey the same kind of authority in pixels, so the task becomes twofold: connecting to the existing credibility through branding, and capturing various kinds of word of mouth.

In the coming months, several trust stories will bear watching:

-Pharmaceutical companies, particularly in the COX-2 (Vioxx) neighborhood, have suffered major reputational damage, much of it related to online behavior, and the legal proceedings will be only one element of a fight to regain public trust.

-The 2008 presidential race will begin heating up, particularly the early-stage fundraising. Watch for the lessons various candidates learned from the Howard Dean experience.

-After the "golden age" of the CEO as hero, the past few years have reversed the public reception of business leaders. Huge severance packages following poor shareholder results, lawsuits, criminal guilty verdicts, and general tarnish on the aura make for a tough time to be a leader. Will Mark Hurd fare better at HP than did Carly Fiorina? Can Ford and GM rise to the challenge of viability and profitability? Will Boeing build a lead on Airbus? In each case, much will hinge on how much trust the leader can generate in his or her own company, the market, and the financial community. So far, by the way, it appears that Hurd understands the power of e-mail better than Harry Stonecipher at Boeing, who apparently let it become his undoing.