Perhaps Microsoft Doesn't Need a Roadmap for Exchange, Dec 17

Microsoft doesn't have a roadmap for its email and calendaring server, Exchange, even though it is one of the two leading products that deliver email capabilities within the enterprise market globally. I have written previously that Microsoft is wrong not to have one, have said as much in consulting sessions with enterprise clients, and a string of recent articles have commented on Microsoft's lack in this area. All of this has resulted in a great volume of silence from Redmond, which has led me to question whether Microsoft actually needs a roadmap for Exchange Server at this time. This piece explains my current thinking, and I invite you to share your thoughts and perspectives below.

Vendors Need Roadmaps
In general, I am firmly of the belief that vendors of enterprise-class software products should have a published product strategy. This is for two main reasons: (1) the implementation of an enterprise software product is costly when undertaken, and incurs a stream of maintenance and administration costs on the organization until the product is replaced. (2) the organization needs a level of comfort around the ongoing fit between a selected product and its technology needs. In other words, if the vendor is planning on changing the product 3 years out to something different, then that change should be disclosed to the organization for use as a data point in making the upfront decision.

My view is that a roadmap should cover 5 years, with three major sections:

• 1 year out ... the features and capabilities that will definitely be added to the product in 1 year. These should be under current development, and nearing beta release.
• 3 years out ... the features and capabilities that the vendor is seriously considering for release in 3 years, and on which exploratory research is currently being undertaken.
• 5 years out ... the potential features and capabilities that could be added to the product, although these only exist as potential ideas and are not attracting any significant level of current effort.

Does Exchange Need a Roadmap?
If we specifically consider Exchange Server, does Microsoft need a roadmap for it at this time? Here's some data points and thoughts:

• 40% of customers using Exchange are still on Version 5.5, which is two major versions behind the current state of play (5.5 --> 2000 --> 2003). 5.5 was released in 1997. Perhaps a roadmap of versions beyond the current 2003 release is unnecessary because so many of the installed base already have an "available roadmap" before them. Eg, next action: upgrade to Exchange 2003.
• Exchange only delivers email and calendaring, and those areas are mature and almost commodity in nature. A roadmap isn't needed for Exchange because there isn't a lot more that could be offered ... just tweaks around the edges, rather than fundamental changes. In the mid and late 90s Exchange was positioned as "Lotus Notes killer", but Microsoft has abandoned that line of approach in Exchange, choosing instead to focus on killing Notes via a collection of servers and services in the wider Office System.
• Microsoft hasn't got its roadmap act together. Important deliverables, such as the new file system called WinFS, were supposed to be delivered in the Longhorn wave of products, but in the last week Microsoft has announced that WinFS won't ship in Longhorn Server until the next decade. If Microsoft is getting fundamentally important deliverables wrong in its "1 year - 3 years - 5 years" plans, then perhaps it is best for it not to extol future ideas re Exchange.

What Do You Think?
Does Microsoft need a roadmap for Exchange Server? Why? Please leave a comment below, or email me at michael.sampson@shared-spaces.com.

Comments

[Peter de Haas]

Michael,

Good approach to this discussion I must say ...
I don't agree with all you comments however :

...40% of customers using Exchange are still on Version 5.5, which is two major versions behind the current state of play (5.5 --> 2000 --> 2003). 5.5 was released in 1997. Perhaps a roadmap of versions beyond the current 2003 release is unnecessary because so many of the installed base already have an "available roadmap" before them. Eg, next action: upgrade to Exchange 2003. ...

The marketshare of Exchange 5.5 differs al lot per region / country. One of the hurdles for companies migrating to newer versions is AD. A lot of companies are quite satisfied with their current Exchange version and do not see the need to upgrade to any new version of any vendor for that matter. The 'Move to Lotus' campaign doesn't get them moving either.
What Exchnage 2003 does offer them is big saving in ROI through server consolidation and better manageability and usability. I do agree that the 2003 version is already a leap into the future for most customers. Especially when you see that in the context of their overall productivty and communication toolset.

...Exchange only delivers email and calendaring, and those areas are mature and almost commodity in nature. A roadmap isn't needed for Exchange because there isn't a lot more that could be offered ... just tweaks around the edges, rather than fundamental changes. In the mid and late 90s Exchange was positioned as "Lotus Notes killer", but Microsoft has abandoned that line of approach in Exchange, choosing instead to focus on killing Notes via a collection of servers and services in the wider Office System...

Finally someone who sees it. The shift is not only in the Exchange productline. The new ballgame is Communication and Collaboration. Email / Calendaring is an important part of that. On the orther hand I do see a shift towards Real Time Communication (IM,VOIP related to Presence). The "collection of Servers" is just a matter of time and very much of perceptions. Count the number of "single Server" components in an average Notes of Wrokplace shop. Maybe better marketing, but the same solution components in general.

...Microsoft hasn't got its roadmap act together. Important deliverables, such as the new file system called WinFS, were supposed to be delivered in the Longhorn wave of products, but in the last week Microsoft has announced that WinFS won't ship in Longhorn Server until the next decade. If Microsoft is getting fundamentally important deliverables wrong in its "1 year - 3 years - 5 years" plans, then perhaps it is best for it not to extol future ideas re Exchange...

Agree that Microsoft often has had challeges with release dates of products. WINFS being late to the party will be missed, but the roadmap is not tied to WINFS alone. A lot of important products will be released within the timeframe. Yukon (SQLSERVER 2005), Whitbey (Visual Studio.NET 2005), BizzTalk Server 2004, Vitual Server 2005 and LCS 2005 / Istanbul. I bet you also Office 12.

In summary, challenges that are not unique to Microsoft. I do agree that roadmaps are an important means of communicating the future direction, but so is a companies current performance, strategy and way of excuting. As you've righfully stated, the current versions of a vendors software solutions are often "the future" for the majority of clients.
Take the example of the IBM. How many Workplace seats have they deplyed within thier own company ?

[Michael Sampson]

Peter, thanks for joining the conversation, and posting your thoughts. My feedback:
- if AD is such a "hurdle", perhaps Microsoft needs to do something to remove the hurdle for those that don't want to embrace it.
- I agree that "a lot of companies are quite satisfied with their current Exchange version" ... that speaks to the fact that even at Version 5.5 the product was fairly mature. It is just email and calendaring, although a case could be made that actually there is less functionality in the more recent versions as capabilities were stripped out and put into different and new servers. If you followed that to its next step, perhaps the next version of Exchange would remove calendaring?
- what sort of ROI savings are you seeing with E2003 migrations? How does that break down between savings on licensing costs and savings on administration and management?
- are you alluding to a future roadmap item with respect to a "collection of servers" and "just a matter of timing and perceptions"? Will the market see a simpler way of embracing Microsoft's communication and collaboration strategy, which currently involves taking on 5-12 servers, depending on how you cut it? Licensing deals with the problem at one level, but the organization still has to do the maintenance and administration over time.
- you commented that a "company's current performance, strategy and way of executing" is an important means of communicating the future direction. Specifically with Exchange, one has to say that MS isn't doing too well on your three attributes: 40% of customers globally are on a version of the product that was released 7 years ago ("performance: poor"), notwithstanding what I've written above, there is no sense of 'what's next' ("strategy: poor"), and those things that were recently talked about and 'promised' have been re-scheduled (Edge Services), canceled (Kodiak), or severely delayed (SQL-based Exchange) ("way of executing: poor"). Generally speaking, however, Microsoft is going to be a long term player whichever way you cut it ... it has (a) a monopoly position in two key markets -- desktop OS and Office Productivity Suites, and (b) a mountain load of cash. However, in my view, it faces ongoing issues with security, loss of customer confidence, and a huge R&D expenditure that isn't returning much (with the Tablet PC being a notable exception).
- IBM? Any response on the last sentence?

[Ed Brill]

I don't think internal deployment of Workplace is a critical metric, when Notes is already 100% deployed at IBM. It's like asking a Ford Executive "Why aren't you driving the new Jaguar?" when he already drives the new Mustang. The equation will change as Notes and the Workplace client technology are brought together; certainly then IBM will be able to demonstrate that no migration is necessary to Workplace, but rather it will be integrated with the current architecture.
Workplace's 1 million existing sold seats are often found in shops running Exchange, who haven't been able to realize any collaborative value from their Microsoft investment.

[Ed Brill]

I also completely don't get this comment from Peter: "Count the number of "single Server" components in an average Notes of Wrokplace shop. Maybe better marketing, but the same solution components in general."

A Notes shop runs a Domino server. There may be some cases where a technology like LEI runs on just one Domino server, but it's still Domino. There aren't 12 servers required just to get the base capabilities.

[Kent Duston]

Michael,

Leaving aside the will-they-or-won't-they strategic direction implications of not having a roadmap, there would seem to be a commercial issue at stake as well.

As we all know, major vendors have major product lines that require multi-year investments, both from them (to design, build and release) and from us (to plan, migrate and deploy). Unless the vendor is in the retail end of the business, the item that keeps revenue flowing while the design-build-release-migrate-deploy cycle works through is the annual software maintenance fee.

Software maintenance seems to me to be one part insurance (I'll have the upgrade when I need it) and one part faith (a - sometimes misplaced - hope in things unseen). While everyone allows for software maintenance in their IT budgets, it's generally with the expectation that it is a way of paying for future upgrades with today's cash. By and large, it's good business for the vendor and a safe play for the customer.

Roadmaps and their related marketing ephemera seem to me to be a key part in extracting that annual maintenance fee. I pay my annual fee in the expectation that I can upgrade in the future - but if there's nothing on the roadmap, how long will I perservere with that behaviour? What's the point of paying my deferred upgrade costs if there's nothing on the horizon to upgrade to?

Of course, the argument goes that you migrate rather than upgrade, but given the very price-attractive bundles that vendors typically use to provide cross-grade incentives, it's probably cheaper to simply ditch the maintenance altogether and pick up a well-priced deal when I'm ready.

As you've pointed out, from a strategic viewpoint there's an argument each way on whether Microsoft really needs a roadmap for Exchange or not. But I can't help thinking that steadily declining software maintenance revenues will be the sting in the tail for remaining silent.

[Peter]

Still not sure where the "12 server" number comes from
To provide full Collaboration / application functionality the following servers are required :
1) Windows Server 2003 (including Windows SharePoint Services)
2) Exchange Server 2003
3) SharePoint Portal Server 2003
4) Live Communication Server 2005

What am I missing ?

Ed : Not every organisation can afford two cars in the garage and plans to mut the Mustang Engine in the Jaguar ;-)

[Ed Brill]

You're missing SQL Server for any real size of deployment. And Live Meeting services for those who want real e-meetings.

[Peter de Haas]

OK, but including SQL I count 5 servers, not 14.
Is it really that more complicated than LN 6.5.x ?

" The entire Lotus Domino-based portfolio of solutions (Notes and Domino, Lotus Instant Messaging and Web Conferencing (Sametime), Team Workplace (QuickPlace), Lotus Workflow, and Domino Document Manager) is now on a common release schedule, with a common set of operating systems, languages and supported browsers."

When you consider the fact that a lot of organisation running Lotus Notes often also run multiple Windows Server machines and often also SQL Server I think it's a lot less complicated all of a sudden...

[Peter de Haas]

oh yeah, LiveMeeting is a service, not a server. But you knew that I assume.
Integrating this with the rest of the Office System environment is the least complex of all I would say.