Continued from page one

What's more, the company keeps celebrating their intelligence. They're hired because they're the "smartest"; they advance because they work "smart," which in the Microsoft lexicon means not just intelligent, but also creative, intuitive, disciplined, market-savvy and persistent.

The smartest man on campus is Gates - a towering but intimate deity, one often invoked in discussions or referenced in the interoffice e-mail that is the Microsoft nervous system.

"No one thinks that they're smarter than Bill," says Peters. "Nobody thinks they could run [Microsoft] better than Bill. It creates order when you don't think that your boss is an idiot."

What also creates order is Microsoft's sense of mission. "Bill and Microsoft believe that any problem can be solved by smart people who are focused and hard-working," says Maples. "People don't mind working 15-hour days. The rest of your life doesn't count for much. If you're one of them, it's perfect." Though Maples no longer works full time at Microsoft, he remains an "ambassador" for the company and counselor to its top executives.

"Money isn't the motivator," says Julie Larson, 36, a FrontPage team manager who worked her way through school, earning a master's degree in software engineering. "Instead, it's the desire to succeed, to succeed with your peers, to do things that change people's lives."

That, plus the gnawing fear that you may not be worthy. "Microsoft plays on your insecurities," Larson says. "How good are you? You're competing with others in the industry, competing internally with others. There's a lot of competition."

An informal after-work retirement party is held for Brendan MacLean, who is leaving Microsoft. (TWP)

Programmers loom large in this competition. It is their skill in "coding," in turning strings of numbers into life-altering software, that is Microsoft's lifeblood. "Programmers are like artists," says Peters. ". . . When you're writing code you're not thinking in English. . . . It's like a play - there's motion, things work, it's not static. You know where you're going. . . . Things just flow."

Programmers don't waste much time thinking about a program's possible bugs. They're more prone to envisioning what the program should do, rather than what it actually does.

On the other hand, they know that the art of writing software involves making compromises. At Microsoft, particularly, there is a sense of urgency about getting the product out to customers. You gotta ship the software. Getting the product into the user's hand is considered an integral part of the program, equal in importance to any given feature in the software. "Shipping," in the words of Chris Peters, "is a feature."

The FrontPage team members discovered this from the start: They faced a tight deadline to produce the next version of their software - and the version after that. "On Day One, you're already overwhelmed with work," says Stefanik. "You're behind, with a huge list of stuff you've got to get done. Training comes from making mistakes."

You gotta ship the software. Shipping is a feature.

Product testing in a Microsoft "usability lab." (TWP)

In April 1996 - just three weeks after the full Vermeer group officially arrived in Redmond - they shipped FrontPage 1.1, with the Microsoft imprimatur.

And within two months, FrontPage sold 75,000 copies. Two months later, its sales had doubled. According to the marketing research firm PC Data, that was more than 30 percent of the retail market for Web publishing tools.

Microsoft's Advantages

In launching FrontPage, Microsoft enjoyed many big-company advantages.

There was the power of its brand name - an asset Microsoft spends about $120 million a year to promote. Customers might be wary of investing in software produced by a start-up, no matter how cool it might be, because the start-up could go out of business overnight. No one has that concern about Microsoft.

There was Microsoft's distribution network. Vermeer's customers had to dial an 800 number to order FrontPage. Now, they could find it at hundreds of retailers.

For a break in routine, software developer Charles Parker takes a few shots at the basket in his office. (TWP)

There was Microsoft's reach. As FrontPage 1.1 was in development, the company worked closely with major Internet service providers to ensure that they could easily "host," or support on the Internet, Web sites built with the software.

There was Microsoft's pricing. Vermeer had offered FrontPage at $695; Microsoft sold it for $149 - and offered rebates to people already using certain Microsoft products.

Software pricing is an arcane science, or maybe an occult practice: Manufacturing is so simple - copy program onto disk; package disk; shrink-wrap and ship - that manufacturing costs hardly enter the equation. (Even easier: Let customers fetch the software from the Web.) Software companies big and small simply try to pinpoint the maximum that the likely audience will pay. And oddly enough, start-up companies often charge more for their software, on the theory that people who take the risk of buying a start-up's software must really want it.

Vermeer had been selling FrontPage to sophisticated users. Microsoft, with its pricing, put FrontPage within easy reach of a much bigger segment of the market - and it was a growing market.

Microsoft's competitors watched with awe and trepidation. Their sales were increasing, too, but not at that pace.

From Microsoft's vantage, the competition was intensifying. Adobe Systems Inc., a leading maker of desktop publishing software, was a player. Netscape was working on Web tools. And start-ups were also diving in, including NetObjects, a company based in Redwood City, Calif., that would emerge as a player in the market for Web tools (and is now half-owned by International Business Machines Corp.).

Spurred by its sense of the competition - and by the opportunity it saw in the Web - Microsoft pushed the FrontPage team to churn out new versions.

In fall 1996 - just six months after the release of FrontPage 1.1 - the team pushed out FrontPage97 in six languages. The feat demanded 80-hour workweeks. By spring 1997, PC Data reported that FrontPage commanded about half the market for Web publishing software.

In fall 1997, Microsoft rolled out FrontPage98, again in six languages. The following February, FrontPage could claim 1.5 million users around the world, and 64 percent of the overall retail market for Web tools. The software had also accumulated 40 awards for excellence from industry trade groups.

Last spring, just as the FrontPage team was limbering up to produce FrontPage 2000, Marketing Manager Pat Kirtland, an energetic 32-year-old, called a meeting to rally his teammates. "There's nowhere else to go but down!" he said.

Around the same time, Microsoft executives were mulling over a strategic question - whether to marry FrontPage with one of Microsoft's core products, its Office suite. Office is a package that combines five of Microsoft's best-selling products - including word processing, spreadsheet and e-mail software - so that they work smoothly together.

At that point, FrontPage had 2 million users. Office had 75 million.

Changing the Product

The FrontPage developers, however, were focused on their product. Whatever its reputation in the outside world, Microsoft's approach to product design can be humbling for its engineers.

A little after 11 on a morning in mid-March, three FrontPagers were slouched in Microsoft's standard- issue red-covered chairs in a windowless conference room, listening to Richard Jacques, who runs "usability tests" in which Microsoft invites consumers to try out software in a suite of rooms with one-way windows and loads of recording devices.

"Everybody tried to drag-and-drop, then right-click," he said in his British accent. "I was surprised by that." What he meant was that to move a picture, test subjects tried to use a trick that works in other Microsoft programs - but didn't work with FrontPage.

"That's easy to add," said one team member clad in sweat pants, sweat socks and sandals.

"Well, maybe we need to put this feature somewhere where it's more discoverable?" Jacques said.

This exchange ignited a 45-minute skirmish over where most users would expect to find the option for moving a picture. The most respected voice in the debate belonged to 25-year-old manager Mike Angiulo, who suggested that the FrontPage team take an example from Microsoft's Excel program. Outside the conference room, the halls were quiet except for the swishing of sneakers padding down carpeted corridors.

As the weeks went on, another issue compounded matters: The FrontPage crew had to wrestle with how other Microsoft products worked. How would FrontPage have to be changed as it was drawn further into the Microsoft family?

When FrontPage managers met with some counterparts from the Access team in mid-April, the gathering had the polite chill of a meeting of in-laws.

Access, a database program, relies on a different core technology from FrontPage; there were, in Julie Larson's opinion, enough subtle differences between the two programs to confuse customers who used them both. Neither group wanted to change its approach, but one ultimately might have to: For customers, consistency counts.

The issue is complicated by the fact that an Access manager received "billmail," or a short edict from Gates. It said, "Don't use [a particular technology] in this anywhere," said Angiulo.

How Access will rejigger its code and what it all means for FrontPage isn't clear. Top executives "will give us the 10,000-foot view," Angiulo said, then back off, letting the groups hash out the details. It takes time.

In dozens of meetings over the next few weeks, Larson and Angiulo played through similar scenarios with other groups. FrontPage already had been changed so that it looked and worked like other Microsoft products. Sometimes, FrontPage got a boost by taking advantage of such work - but sometimes it didn't.

"Sure it would be easier to get FrontPage out the door . . . if we just focused on FrontPage," Larson said. "But in the long run, it's better."

After 2½ years within Microsoft, the FrontPage group was still a novice at working with other groups. To Peters, FrontPage seemed to have much of the old independence that all Microsoft groups once enjoyed. Customers may appreciate such tight integration, he said, but "it adds complexity."

And if Microsoft were to make some dramatic marketing decision - say, to package FrontPage as part of Office - it would add layers of complexity the FrontPage team hadn't had time to imagine.

Concerns About Packaging

Office is a powerhouse - almost as integral to Microsoft as Windows. Last year, it brought in even more revenue than Windows - about 40 percent of Microsoft's $14.48 billion in revenue, according to analysts at investment firm BT Alex. Brown Inc.

Packaging FrontPage with Office would give Microsoft advantages beyond expanding FrontPage's potential market to 75 million users.

Continued on page three