Listen

Description

Each module of Office deserves a shot at being the hero of a release. That’s how a healthy product bundle should move forward, rather than relying on a single anchor. Excel 5.0, Word 97, PowerPoint 2000, Access 2.0 anchored Office Professional. While Outlook was top of mind for IT infrastructure managers, it remained complicated and frustrating for regular end-users. Embarking on a radical redesign of a major product is a career opportunity and also a big bet for a critical business that continued to be half of Microsoft. That would be difficult enough, but nothing is that straightforward. Outlook would still face internal pressures of strategy and alignment that have made building a breakthrough release so difficult previously. Would this be the moment for Outlook to shine?

Back to 073. **DO NOT FORWARD**

Outlook barely worked. Still.

Such was the product-market fit of Outlook that enterprise customers owning the latest in all the Office tools were routinely deploying the newest Outlook while leaving old versions of the core Office apps on the PC.

Five years and three releases from the debut of the product, Outlook remained fragile, bloated, and too difficult to use. It wasn’t as though the team hadn’t executed, releasing Outlook 97, Outlook 98, and Outlook 2000 in the span of just over three years. Rather the team, through little fault of their own at least after Office 97, whipsawed through strategic initiatives. First, they had to split the product into internet and corporate modes. Then they had to merge the product back together while attempting to integrate with Office and the deployment tools of Office 2000. Then they had to charge up the hill of unified storage for the second time, only to cut the feature again at the tail end of the project. As if this wasn’t enough, the past couple of years saw the rise and unveiling of NetDocs which pivoted to a potential Outlook replacement, only to see that vision meet reality.

Marc Andreessen, founder of Netscape, wrote in 2007, years and a generation after the release of Outlook and long after the demise of Netscape, “The Only Thing That Matters”. This short piece codified product-market fit. Essentially, he said that the market pulls a successful product out of the company, “[t]he market needs to be fulfilled and the market will be fulfilled, by the first viable product that comes along.” Outlook was such a product. Coupled with Exchange the market simply needed the combination to exist and to be provided by Microsoft. No matter what, the market was going to make the offering successful. It simply didn’t matter what Microsoft did or did not do, Exchange and Outlook were going to be successful. Once customers saw enterprise-grade email running on Windows Server with a single integrated mail and calendaring solution, all from Microsoft, everything else was set aside: bugs, bad user-interface, poor performance, and missing features. The combination clobbered IBM Lotus Notes. The rest is history.

We (or I) could not overcome the internal forces driving Outlook’s strategy in order to give the team space to focus on building the product it needed to be. It was weird to achieve so much success with such a challenging product. That really throws one off balance. Product people, so to speak, like to think that being a great product is what matters in the market. This belief drove so much of the dialog internally and across teams it is hard to think of any other factors that were ever discussed—the battles were over what makes for a great product. We battled features, architecture, performance, competition, and more. We never really debated the other aspects of the 4 P’s of the marketing mix beyond product: price, place, and promotion. Outlook was the right price, available through the right channel, meeting an articulated need exactly the right way. Oh, and Outlook email and calendaring kind-of-sort-of worked.

Still as a product development organization and leader, I really wanted to make Outlook work. We really wanted a release of Outlook we could be proud of in a product sense. It was a mission.

Standing in the way, more strategy.

The stars were aligning to allow us to focus—Exchange 2000 shipped and was stable, and Windows simply moved on from Outlook Express to aim for a grander and more unified Longhorn plan, the successor to Windows XP in the early stages as of this time. Yet, we found ourselves in the middle of a Hailstorm.

Hailstorm was the code name for a set of services announced shortly after Forum 2000 that built on the announced .NET strategy. Hailstorm would provide email, messaging, notifications, storage, and more, aiming to be the foundation for a broad set of consumer internet services.

That hardly stopped the corporate strategy motions. No sooner had Office XP shipped than the Hailstorm juggernaut took aim at Outlook. Once again, the classic Microsoft platform play entered the picture—a platform needed apps, and apps needed to demonstrate the value of the platform as the first and best adopters. Plus there was that other part of the strategy play, which was that the platform wasn’t even close to complete. Pushing apps to support a new platform sooner accelerated its completion. Such a dynamic worked once, with Excel for Windows 2.0, but Windows was under development for several years before Excel really began to drive the platform with feedback. Plus, Excel had already made GUI work on Macintosh. This approach did not work with OS/2, though maybe IBM shouldered some of that blame. It could be said that Outlook contributed to Exchange success, but that was not a feeling shared across teams who did not see Exchange success as a client as much as success of a server (I disagree). Hailstorm managed to get a mandate for Outlook support, old-school Windows-Excel style.

Outlook was going to get beaten with the strategy stick once again. There was, however, one problem.

The Hailstorm platform was marginally more than memos and specifications, and few understood how it would diverge from Exchange (or Hotmail)—in other words Hailstorm implied supporting a new mail protocol for a client that only understood Exchange and was still not very good at internet standards. There remained an old-school belief that the world would accept a new proprietary mail protocol for internet scale email. Hailstorm was going to build on Hotmail infrastructure, but somehow introduce a new scale platform. Underlying this was the assumption that Outlook was architected to accept plugins supporting any email protocol so long as some small amount of protocol-specific code was written. This classic Microsoft miscalculation on the utility and viability of such architectures caused many cross-group skirmishes and much finger-pointing.

What many never really understood about Microsoft’s email strategy and execution was that Outlook and Exchange were tightly coupled, just as Lotus Notes and later Gmail were for their respective clients and servers. They are essentially one product built by two teams. Even though Outlook could sort of support other mail servers, such as Yahoo or Hotmail, using those was never as good as when using Exchange—many features in Outlook required Exchange, with users mostly shouldering the burden of figuring out which features worked or not. Importantly, vast numbers of features were not expressed in the industry standard email protocols that existed and exist today. Hailstorm tried to build a mail server without a client. Supporting Hailstorm amounted to rewriting Outlook to support whatever it was that Hailstorm decided to implement as email, calendaring, contacts, and more.

What became clear was that the three-year Enterprise Agreement train that the Office11 schedule demanded would not be enough time for Hailstorm to deliver a working, solid, and scalable platform for Outlook. Hailstorm needed more time. Beyond Outlook even, Hailstorm proved to be too much, too soon for the many partners it needed.

The world of co…



This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit hardcoresoftware.learningbyshipping.com