The suite squeeze.

Dear Soft*letter,

Just back from Comdex. Don't know if there was an official show "theme," but the big splash was certainly the advertising war between Microsoft Office and Lotus SmartSuite. (One example: Giant neon-yellow balloons in the Convention Center parking lot, painted to lock like SmartSuite boxes. Hard to overlook, even against a Las Vegas landscape of flaming volcanoes, black glass pyramids, and Arthurian castles.) Which brings me to my question: Why are these guys spending so much cash to promote software they're practically giving away? Do they know something the rest of us haven't figured out yet? Your thoughts, please.

Mixed-up in Missoula

Dear Mixed-Up,

Yes, the suite wars have now become the high-stakes game in the software marketplace. What started as a price-cutting gimmick has quietly turned into a very different kind of competition based on software integration issues. Suites are no longer just a collection of individual applications; they are evolving rapidly into fully integrated environments, complete with their own interface conventions and rich collections of shared features and services.

And it's not surprising why Microsoft and Lotus both see suites as an do-or-die market opportunity. Rather than pick "best of breed" applications from a variety of developers, suite customers tend to lock themselves into a vendor's complete product line, with a single tech support connection, common end-user training, and (eventually) a common upgrade cycle. The potential for long-term customer control goes far beyond anything the industry has seen before.

And it's clear that users are increasingly enthusiastic about this trend. The market research numbers are a bit murky, but suites now probably account for at least a quarter of total Windows desktop applications sales in the U.S. (higher in Europe), with a growth rate in the 60% per year range--higher than in almost any other major Windows category. Microsoft, with more than a billion dollars in annual suite revenue, is the clear front-runner so far in market share. But this time Lotus hasn't wasted time fumbling at the starting gate: The SmartSuite line is uniformly strong and Lotus's marketing has been right on target.

Regardless of whether Microsoft or Lotus comes out on top in the market share race, however, it's increasingly obvious that the suite model represents a serious threat to just about everyone else who competes in major productivity software markets. Single-title developers like Software Publishing Corp. already find themselves squeezed out of key accounts; Borland and WordPerfect, who have tried to catch up by co-marketing a suite of loosely-connected titles, don't seem to be much better off.

If this trend continues--and we think it will--one plausible outcome is that the Microsoft Office and Lotus SmartSuite will turn into what are essentially operating environments. Both suites already offer very rich layers of shared services that are accessible only from within suite components. Microsoft Office, for instance, now includes a common scripting language, spelling checker, end-user dictionary, graphing and querying tools, IntelliSense help technology, font effects utility, program launcher, and an assortment of graphics and text filters. For Lotus SmartSuite, an equally important integration turbocharger is interoperability with Notes, which gives SmartSuite applications an edge in workgroup functionality.

Meanwhile, both Microsoft and Lotus have put a good deal of work into designing suite-specific interfaces that go well beyond the limited conventions of the basic Windows environment. Microsoft's Pete Higgins told us recently about the "pixel police" at Microsoft, who make sure every product group conforms exactly to standard specifications for toolbar layout, icon design, menu names, and other interface elements. The work of the pixel police is especially visible (or invisible) when a user moves from one OLE-embedded application to another: Except for a few application-specific icons and menu titles, the screen literally doesn't flicker by a single pixel. "You may not even notice when you've changed programs," Microsoft brags in its Office ads. "That's the whole idea."

In fact, OLE provides one of the most compelling reasons for all Windows developers--not just Microsoft's product groups--to embrace pixel-police rules. In theory, OLE (and Apple's rival OpenDoc standard) gives users transparent access to multiple applications that have been "embedded" within a single document or form. Simply by moving the cursor into an OLE drawing or table in the shell document, the user activates editing tools from a separate drawing program or spreadsheet.

Trouble is, nothing in OLE clearly warns users that familiar keystroke commands and interface conventions may change whenever the cursor moves into a new application. As a result, the embedded applications  (and the shell application) seem to behave in quirky, unpredictable ways that violate all the common-sense rules of usability.

To cope with this problem, at least one developer, Shapeware, has already adopted pixel-level conformity to Microsoft's Office interface.

When Shapeware's Visio 2.0 drawing program pops up inside Word Or Excel, the program feels very much like a natural extension of the Office environment, and Visio drawings behave as if they're a fully integrated part of the underlying document.

The downside to this approach, of course, is that an Office-compliant application like Visio will feel equally unfamiliar when it pops up inside a SmartSuite application. And that's the crux of the problem that suite-based interface standards pose for third-party developers. The more tightly a third-party product is integrated with one suite, the more awkward it will seem in a competing suite environment.

This point hasn't been lost on the major suite vendors, who are beginning to solicit more third-party support for their suite-based interface standards. (In an interesting turnabout on the look-and-feel issue, Lotus now encourages outside developers to license its SmartSuite icon designs.) In the same vein, we suspect that the majority of corporate developers will come to model their own application interfaces on the standards that suites define, and even companies whose products don't mesh closely with suites--for instance, consumer titles--will probably end up adopting much of the standard appearance and behavior of the two dominant suites. Rather than compete head-to-head with Lotus and Microsoft, a lot of single-title publishers (and the legendary software "component" developers) will almost certainly choose to make their products fit as seamlessly as possible into Microsoft Office or Lotus SmartSuite.

But that still leaves most developers with the perennial question of which environment to support. The choice, as always, won't be easy: Microsoft Office is likely to remain the most popular suite environment in the mainstream desktop market. But at the same time, Lotus is bound to leverage its Notes links into a stronger SmartSuite presence in high-end networking markets (where Lotus is now also the only company with a full OS/2 suite). The choice of suite standards will be further complicated by all the usual unknowns--corporate mergers, new and emerging technologies, choices of development tools, and the like.

But if suites do continue to evolve into increasingly tightly integrated environments, staying neutral is going to be tough (and risky). If the Windows revolution teaches any lesson at all, it's that the worst possible development strategy is to wait for the market to set a standard and then race desperately to catch up.