What happens when your biggest strength and greatest asset as a product development organization becomes your biggest weakness? Perhaps that is inevitable. With so many potential disruptive forces, at least that’s what we were hearing, it was almost too much that the very thing we really excelled at—building new features—would become a problem. While the idea of software bloat or even the phrase bloatware was hardly new, it was being applied increasingly to Office. This is the story of the first attempt at doing something about this issue.
Note to readers: Substack introduced a new feature this week—free excerpts with the remainder of the post for subscribers. I’m trying this out to see what subscribers and potential subscribers think. I do welcome feedback. The proceeds of this work do not benefit me, and subscribers can also join in the comments and discussions.
Back to 051. HTML: Opportunity, Disruption, or Wedge
Features. That’s all we ever talked about. Adding features. Fixing features. Missing features to be added or fixed. Office achieved the position it achieved, as precarious as it seemed now with the rise of the WWW and Internet, by adding more features with more regular releases of new versions than competitors. Reviews focused on features and we won reviews. We had built a team, an engine to add features.
As fast as Clippy could appear after hitting the F1 key, it seemed as though our greatest strength and our most significant asset—features—had become a great weakness. We went from dominating with more well-executed features to being crushed by the perception of the weight of our products. The industry latched on to the expression bloatware to describe products that seemed to have too much—too many features, too many megabytes, too slow, too difficult to use, or simply too much.
It was one thing to develop a strategy to address a customer problem we had created, albeit inadvertently, when it came to total cost of ownership. We even committed to building fewer productivity features, despite the internal backlash over personal productivity moving to priority six. It was entirely another thing, however, to look at what we built and be pressured into admitting customers didn’t want or need it. Still, customers were buying Office by the tens of millions of copies.
Were customers, press, and analysts right? Did we have a product problem, a technology issue, or a marketing challenge, or some combination? Of course, the development team was certain marketing wasn’t convincing people how amazing the product was. Marketing was certain the dev team was not delivering business value. Press and analysts were relentless. What could we do?
Our hometown Seattle Times reporter, Paul Andrews wrote “already some have nominated Office for Bloatware of the Year.” While unclear who “some” were and knowing there was no actual award (thankfully), that was a sharp dig in an otherwise positive Office 97 review.
Worse, however, was the headline in the Wall Street Journal that simply stated “Microsoft May Face Backlash Against ‘Bloatware’” right on the front of the Marketplace section. The article used every possible way to explain the scale of Office as big from “two years and several hundred million dollars [in R&D]” to “appetite of companies for such programs” it did not let up. The requisite quote at the start of the article from an IT professional was brutal “couldn’t care less… there was nothing from a business point of view that was a compelling reason to upgrade.” To this day I still get a sick feeling in my stomach from the box in the story stating, “Microsoft’s Office 97 contains 4,500 commands for features both useful and arcane.” Each of those features meant something we were so proud of and represented real effort. Plus, during the interview the reporter kept pushing for ways to talk about how big Office 97 was. I resisted and let this slip in a moment of pride, only for it to be used against me. I added this article to the binder I carried around and had it at the ready whenever the topic of bloatware came up.
Bloat was a constant source of strain in conversations with field sales, in the Executive Briefing Center, and with the press. Each time the topic came up I used my go to answer, which was that there is a set of features that everyone in Office uses and those are easily understood such as open, save, print, copy and paste, and a host of basic formatting commands. Everyone nods. Then I would talk about how each application has features that a few people use, perhaps footnotes, financial formulas, or animations in PowerPoint. Most would agree with that.
A common variant of this discussion was “your bloat is my crucial feature” or “yes this is bloat, except that one time a year I need to use it.” Most PowerPoint presentations are minimal when it comes to production values, except for that one time a year when the stakes are high and huge effort goes into making a great show. A favorite example is that few claimed to use Mail Merge in Word, until they find out they need to send out holiday cards or invitations to a large group (and yes, it even works with email!). Bloat rarely considered the frequency of use and often presumed infrequent use meant no use.
I would point out the significant business value of Office was that any person could use their set of features in any one tool and seamlessly share files and collaborate with someone way more advanced than they were. For example, I don’t know how to draft a contract and use “red lining”, but a lawyer could use those tools and send me the contract for review. Intellectually people understood, but almost always would shrug and still say that the software is “bloated”.
To each constituency, bloat implied something different. There were mundane answers like how much disk space or RAM Office took up. These Moore’s Law measures were easy to complain about but were especially incorrect relative to competitive products, Office was far and away the best. Unfortunately, no one ever experienced more than the one product they used so if Office seemed slow and all conventional wisdom was about needing to upgrade hardware with more memory or a laptop that ran out of disk space, then bloated Office was to blame.
Some believed that literally having too much stuff on the screen is what made the software slower or bloated. There was some truth to this in that the user interface of Office—the menus, toolbars, wizards, and more—was rapidly growing and exceeding the available pixels on the most common screen sizes, especially on new laptops. While desktop computers were getting larger monitors, laptops, still uncommon, were also a generation behind in the amount they could fit on the screen.
The industry analysts believed in something of a combination of all these factors. There was a view that the older legacy features of the product were weighing the programs down with old code, or cruft, and that if the product could just be “factored” [a technical programming word, to break the code up into smaller pieces] then we would have much sleeker and more tuned products that took less memory, less screen real estate. Analysts viewed the disruptive technologies of Java, HTML, and components discussed previously as somewhat magical answers to bloat. Frustratingly, many thought…