The Sirana Roadmap for Exchange

Sigh.

This Exchange Roadmap issue is really dragging on me.  I don't know why.  I guess it's hard to see a once-invincible brand thrash around like this.  Maybe I'll feel better soon; Microsoft is going to begin briefing analysts later this month about their plans.

While we're waiting, I'll give you my thoughts right now, right here, about what Microsoft should be doing.

When you talk about a roadmap, you have to divide it into some short-term goals and a longer-term vision.  The short term stuff you can deliver in the next year or three, and the longer term stuff might take five or more. So let's dive in and take on the short term first.

The Short Term

You frequently hear people asking about Exchange, "It's such a mature platform for email. What else could you do?"  Well, it turns out there's lots you can do, and that's an excellent place to start for our short term vision.

You can start with some easy ones.  Read the newsgroups.  How long have people complained about not being able to do stuff like customize the warning message users get when their mailbox is too full?  Administrators should be able to customize it include links to sites on their intranet, tips, etc.  Whatever.

Think this is meaningless, small-time stuff with no impact?  No.  It's called "Delighting your customer."  It's always a good strategy to say, "We're going to spend 50% of our time on this release to address the top N things that customers hate or want. 

That guarantees that you'll be able to show value to those customers, and hopefully drive their upgrade plans.  How many customers told Microsoft in the late 1990's, "I absolutely must have you integrate the Exchange directory into AD"?  Probably one. Maybe two. It may have been the right strategic move to drive AD deployments, but it's not customer delight.

Sometimes, I think Microsoft builds it's product plans based on whomever showed up in the Microsoft Executive Briefing Center last year.  At the EBC, program managers hear from extremely large customers about their problems.

Unfortunately, those customers tend to skew to the extremely large enterprises who merit an invitation to the EBC.  And also just as often, the attendees are not hands-on administrators.  They are CIOs, or "Enterprise Architects".  I'm pretty sure this must have been the reason why Exchange Edge Services has now morphed into a "content platform" that supports message archiving and other compliance issues, rather than a pure-play security solution. That's top-down thinking.

Is there more in the short term?  Absolutely.  The spam/email security problem is issue #1 with administrators.  If the next version of Exchange solved it (and I'm talking orders of magnitude better than Intelligent Message Filter), customers would upgrade.  In a heartbeat.

When you talk with Microsoft, they go on and on about how good IMF is.  And it's a pretty decent anti-spam engine.  But the best part about IMF isn't it's performance, it's the price.  It's free.  For many customers, free is very compelling.

But if you want to look at the best there is in email security, look at commercial products like Brightmail. Or CloudMark. It turns out it's not just the anti-spam engine that's important.  It's all of the other policy stuff around the engine, as well as the administration, management and reporting tools.

IMF is just an engine.  And it's not entirely customizable.  So even though customers have to pay for commericial solutions like Brightmail, it's money well spent, considering the magnitude of the spam problem.

I know Microsoft often tries hard to support a viable third party vendor market place.  But I really don't care to see a viable third party market place in the spam/content security business.  The existence of this market just provides evidence that the base platform itself (email), is really really sick. And if the base host is sick, that's bad for business.

Maybe Edge Services will be the answer.  But I'm guessing not.  So I think Microsoft should just go out and buy a really, really good antispam/security company, and throw it in the Exchange box.  Deeper integration can come long term.  Although probably 65%+ of customers already have an anti-spam solution, many continually look at alternatives.  And this move says clearly: "Exchange is all about message hygiene and security and we take it very seriously.  We're here to protect you."

Finally, I know I'm going long here, especially in the short term area.  Maybe we're already past that time frame and into long-term anyway.

But for many customers, disaster recovery, or even casual recovery, are still big big issues.  Disaster recovery is when you have some major catastrophy.  Maybe your server dies because of hardware failure and you have to rebuild it.  Or you have a natual disaster.  Casual recovery I think is more prevalent, and that's basically "How do I restore the email sent by my boss last week, that's been deleted?"

Customers have been complaining about this issue forever.  Initially Microsoft came up with a partial solution, called Deleted Item Recovery.  Deleted Item Recovery allows customers to restore mail, but only if they enable DIR, and only if the message is still in the deleted item cache.  Outside of that, recovering an item as small as a single message required restoring an entire server from tape backup.

Later, in Exchange 2003, Microsoft introduced Recovery Storage Groups and Snapshot Backups.  Those help, although you still can't take a Snapshot Backup with the Windows 2003 Backup applet.

In terms of granular backup, the Exchange team has always said "We backup the database by pages for performance reasons, and we can't give impart the logical structure of the Exchange store (mailboxes, folders, messages, etc.) because it would take too long."  That's a fine response, but that's a technical response.  It doesn't address the market needs expressed by your customers.

Vendors like Ontrack seem to be on the right approach with tools that let customers restore individual messages from a variety of sources.  Other vendors, particularly backup vendors, have solutions here also. I think tools from Microsoft for disaster recovery, and more importantly casual recovery still have a long way to go in terms of answering customer needs.

In this post, I've really only talked about three main areas: addressing small areas for customer delight, email security, and disaster recovery.  There are obviously areas that I've left out.  Feel free to point them out in comments.

I'm going to talk about what Microsoft should be doing in the long term in my next post.  That post should be significantly shorter, since Long Term Visions are by nature much less specific.  But there's still much to do.

Comments

[Michael Sampson]

Chris, see here http://www.shared-spaces.com/blog/2005/01/response_to_the.html for my response.

[Dmitry Sotnikov]

Quest Recovery Manager for Exchange is another 3rd party product that allows individual, message-level items, including email messages, appointments, tasks, contacts, and attachments to be found and restored from regular Exchange backups and un-mounted .edb information stores.