If You Disclose Your Roadmap, Don’t Do It the Microsoft Vista Way

by Bill Bliss on March 19, 2011

in Alleged Wisdom,Managing Uncertainty and Risk

There was an interesting flurry of blog posts and tweets the week of March 14th, 2011 about whether it was wise for a SaaS company to publically disclose its product roadmap to current and future customers. Duane Jackson, CEO of the UK-based SaaS accounting company KashFlow, explains why they don’t. Dennis Howlett strongly disagreed with Duane and later clarified his point of view; at this point Ben Kepes chimed in, to which Dennis also respondedAdrian Pearson also made a thoughtful post on the subject, as did Bob Warfield.

It’s interesting reading, but I’ve got a bit of a different take on it. Personally, I believe if you’re asking your customers to bet part of their business on you, some of them will find that pretty difficult without knowing where you’re going. The bigger the customer and the more mission critical the part(s) of the business you enable or support, the more it’s likely to be an issue. On the other hand, if you’re doing great from a sales and retention standpoint without disclosing your roadmap, why start now? It definitely can make execution trickier.

I also agree with Dennis: whether it’s a SaaS product is not particularly relevant. While one might argue that because it’s SaaS the switching costs are lower, that just steers the conversation in a direction you don’t want to go: “We won’t tell you about our product roadmap, but because it’s SaaS, it’s easy to switch to a competitor who will.”

Here’s my take.  I don’t think it’s particularly useful to consider “Should a company publicly disclose its roadmap to customers?” a yes/no question. Only you can decide whether doing so is necessary to attract and retain your most important customers. And the decision may, in part, be forced by your competition.

A more useful question is, if you publicly disclose your product roadmap, what do you disclose, and when? And whatever you do, don’t do what Microsoft did with Vista.

A Brief History

Microsoft started talking about Vista (originally codenamed Longhorn) in July 2001, about three months before Windows XP was released, and released it in November 2006, over five years later. Here were some of the big bets in Longhorn/Vista:

  • WinFS (Windows Future Storage). Bill Gates long believed that SQL databases and filesystems should be unified and made rich enough that applications such as Windows Media Player and Outlook Express could use a common data store instead of writing their own.  WinFS, which started from the SQL Server codebase, was designed to be this common data storage engine. WinFS was cut from Longhorn/Vista and was never released.
  • Avalon (Windows Presentation Framework, or WPF). WPF is a subsystem for rendering graphical user interfaces. WPF did ship with Vista, but it wasn’t stable early enough for any significant built-in Vista features to utilize it. It was originally envisioned as something that would effectively replace the older GDI layer, but it has not done so.
  • .NET Everywhere, including the Windows shell. Although the .NET Framework was included with Windows Server 2003, relatively few applications used it, in part because such applications were slow to load and used a great deal of memory. The idea was that this problem would be solved if the Windows shell UI and other bundled applications were re-written to use .NET. For various reasons, this never happened. (My last team at Microsoft wrote a lot of .NET code that was ultimately thrown away as a result.)
  • Palladium (Next-Generation Secure Computing Base). Parts of Palladium were included in Vista but most of it was not, and Microsoft hasn’t discussed it since 2004.

All four “big bets” were in what I call the “red quadrant of uncertainty” as described in my articles: Overview of the Roadmap Integrity Process and Managing Uncertainty and Execution Risk:

The problem wasn’t that these initiatives were in the red quadrant.  Big bets usually are.  It’s that Microsoft talked about them publicly while they were still in the red quadrant…in some cases before they even knew whether they were truly good ideas.

How Office Did Things

Contrast this to the way things worked in Microsoft Office.  Back in the mid-90′s when I led the product management team for Microsoft Outlook, we used to shake our heads at our colleagues in what was then called the Systems Division. In our opinion, they regularly made their lives miserable by publicly disclosing the Microsoft Windows roadmaps years in advance. Microsoft had a terrible reputation for peddling vaporware, and this was one of the main reasons.

In contrast, the Office Division was far more disciplined about releasing every two years or so, and shared far fewer details about an upcoming release outside the company. (It’s not that we were all so better at creating software; after all, we worked on Outlook for 4.5 years before finally releasing it in the fall of 1996—and took a lot of heat for taking so long. There were a lot of Outlook features that were in the red quadrant for quite some time.) But since everyone in Office knew there was a train leaving approximately every two years, for existing products like Word and Excel we rarely started things we couldn’t finish in that timeframe, and we didn’t tell anyone what was on the train until we knew it would arrive.

How? We used the standard techniques described in Moving Bubbles Towards the Green. And when we finally did publicly disclose features in the next version, they were always in the top two quadrants.

Is it any surprise that since Office veterans Steven Sinofsky and Jon DeVaan were put in charge of Windows, you haven’t heard much about Windows features more than a year or so before release?

The takeaway? Only you can decide whether your company must publicly disclose your product roadmap. But never disclose anything before you know what it really is, whether it works, whether customers even need it, or when you can deliver it.

Tweet this: @bill_bliss New post: If You Disclose Your Roadmap, Don’t Do It the Microsoft Vista Way http://wp.me/p1pkik-5C #prodmgmt

{ 2 comments… read them below or add one }

James March 22, 2011 at 9:45 pm

Bill, great post. Appreciate the simple and sensible approach to roadmap disclosures. What a contrast to our experiences working in the start-up environment! Given your experience in both early stage and mature companies, do you believe there is varying degrees of guidance you would give a company? Or does it hold constant regardless of the size, maturity, and customer base of a given company?


Bill Bliss March 23, 2011 at 3:30 am

Hello James, thanks for reading and commenting.

In general I’d say as little as possible, as late as possible: ideally nothing if you can get away with it. And whatever you say should be as high level and non-specific as possible.

Consider that the ideal; one end of the spectrum. That said, I think that ideal is difficult as a practical matter, as a B2B company at least. At the very earliest stages, it might be; after all few companies are willing to make mission-critical bets on early stage startups; that’s one of the things that makes life in early stage startups challenging!

Assuming you’re past that stage though, I think the best thing to do is take your guidance from your customers and your competitors. By the way, customers should be plural – if it’s just a few squeaky wheels, that alone probably isn’t enough in terms of losing flexibility.

Once you get to the point where you have large numbers of customers and perhaps an ecosystem of developers and/or partners, that’s when disclosure becomes a fact of life.


Cancel reply

Leave a Comment

Previous post:

Next post: