Here’s a question:  Where do you host your email?

Gersham and I are rather well-known for a business we started in 2002 called Geekmail.  By 2003, we were on the cutting-edge of IMAP-based email hosting and ran thousands of mail accounts on a cluster of 9 servers hosted at Peer1 Network in Vancouver.  We pushed the envelope in anti-spam technologies: combining advanced whitelisting techniques with behavioural, bayesian and heuristic anti-spam technologies and using our own common-sense approaches to deliver very high anti-spam effectiveness with a too-low-to-track false positive rate.

What we achieved, in essence, was a sort of email nirvana.  In those days, giving someone 1GB to store their email on your server was unheard of… but we did it.  Hosting catch-all email accounts was a novel concept … but we did it.  Hosting custom email domains was tough stuff too, even, but we did it.  We also had a hell of a launch party. 

A couple of things conspired to force us to close Geekmail… a situation which I will always regret:  1)  We were taken to court by a fool fellow whom we’d (our mistake) taken on as a business partner, and whose sole objective was to kill the company; and 2)  Google launched Gmail.  The latter was far more significant since it was A)  Free and B)  From the web’s hottest property.

Now, this all is the long way of explaining that I am perhaps something of an email geek.  I’ve used one form of computer-based electronic mail or another since 1985.  I co-founded Geekmail, of course, and also did a considerable amount of strategic work for FrontBridge — the world’s #2 message management service provider before its acquisition by Microsoft in 2005.  BuzMe and RingCentral, two Unified Communications services I helped bring to market, were among the first to deliver voicemail to their users via email (believe me, a novel concept in 1999/2000).

Be that as it may, it rather surprises me that even today GMail (which has been in Beta for 5 years) still pales in a number of key features (including anti-spam) to the technologies and quality of services we provided with Geekmail.  While we didn’t have nearly the scaling issues that Gmail has to deal with (except for in our very early days) we never experienced the kind of multi-hour outages that Gmail regularly hands to its users.  We also focused the users’ experience on Secure IMAP, not a web-based interface (though we had one of those too) and offered lots and lots of storage to go with it.  And in our later version of Geekmail, the anti-spam functions were tweakable: if you didn’t like the default settings, you could turn on and off different techniques that were used to combat spam on your inbox.

When we were forced to let Geekmail die a rapid death, we scrambled around to find a company who could take our subscribers and whose service closely mirrored our own.  The short answer was:  there weren’t any.

It wasn’t until a couple of years later that I began talking with John Capo; founder and operator of Tuffmail.  In addition to being a pretty nice guy, John runs a service that is the closest analog to Geekmail that I can find anywhere.  In my view this ranks Tuffmail as the very best email service provider for email geeks anywhere.  And so it has been for about 4 years that I have blissfully run my personal email address at ianbell.com on this service — and am now at a point where it is so critical to how I do my daily business and live my life that I would be miserable without it.

As these screen shots should reveal, Tuffmail is literally like having your own mail server cluster up in the sky somewhere.  By that I mean practically every aspect of its functioning is customizable to your whims and needs.  I can change how it responds to spam, I can block certain servers from sending me mail, I can blacklist any email sender from connecting to the server, and so much more.  I can also have a catch-all, which many email hosts hate to do because it creates spam honeypots, but which has become a critical means for managing my accounts online.

I don’t do any filtering or routing of email at the client level.  This would be impossible, since I access email from four different devices on a day-to-day basis.  Instead, I have input a complex set of rules into Tuffmail’s extremely robust email rules interface (sorry I can’t show you this — classified!) and all incoming mail is stored in the appropriate folder when I check it from my MacBook Pro, my iPhone, my Mac Pro, or whatever.  Microsoft Exchange, Gmail, yahOo! Mail, your ISP’s Mail Server — they all wither by comparison because they don’t allow this sort of granularity — and because they don’t fully embrace standards-based IMAP email messaging.

I keep all my mail, as well, nearly 5GB at the moment.  So if you said something in 2005, it’s pretty easy for me to find that message in my email clients (this could be the reason why mail.app sucks up most of the free memory on my MacBook Pro) and regurgitate it.  This is extremely handy and it reaffirms email’s rightful place at the fulcrum of my life (sad but true).  This is only possible because I have an enlightened email hosting provider who A) embraces large mailboxes and B) embraces large message sizes, which means I can send around presentations and big graphics files without fear of them bouncing back (unless the receiver’s mail server is a dunce, of course).

I don’t ever receive spam in my InBox anymore, because I have the settings and filters perfectly tweaked to my needs on Tuffmail.  But blocking spam is easy these days.  The real problem is blocking it without also blocking legitimate messages — this is much much harder.  And this is where GMail, which uses the Postini service (which is not directly integrated to GMail), tends to fall over.

Have you ever heard the excuse “Oh, I sent you the email, but maybe it got caught in your spam filter….” before?  Sure you have.  That doesn’t happen to me.  The benefit of the Realtime Reports (screen shot above) is that I can go in to the server logs  and actually see when a message flew through or was rejected by the Tuffmail server hierarchy.  I just view the page, do a Firefox search for the person’s email address, and if they sent a message it’ll be there.  I’ve caught anyone who’s ever made that excuse to me in a white lie… not that I hold it against them.  🙂

There is one downside to all of this, of course… with Tuffmail, I have created a monster.  I have so many settings and tweaks, and I have the spam filters so well-trained, that the pain of moving to another provider would be excruciating.  Most likely, I never will.

I don’t endorse products very frequently (and I never do it for any sort of remuneration) — but Tuffmail is one of those rare birds that truly deserves the kudos.  Email hosting is a tough business and in many ways I’m glad I’m no longer in it.  On the other hand, when I use Tuffmail I get pangs of jealousy and nostalgia.  Ah, what could have been!

These prices make Sophos’ acquisition of Vancouver’s Activestate (which, among its products, marketed an interesting anti-spam software solution) for $26M look like a bargain. And with all of these transactions coming in the multiple hundreds of millions of dollars, the number of sellers probably now exceeds the buyers by a 2:1 margin. That means that the remaining two will need to get down to the business of making money.

Fortunately that’s not actually very difficult in the email border security marketplace. The value of anti-spam/anti-virus services for email may soon exceed the money spent on providing email itself. Although the market tracks behind analyst estimates (don’t they all?) email border security was a $3.5Bn market in 2006 and will top $6Bn in 2010, according to Radicati.

The email services business is capital and operationally intensive, and requires a fairly large investment to develop a business. Postini has gone through about $34M in 4 or 5 rounds since its first financing in 2000, from VCs including AltoTech Ventures, LLC; August Capital; Mobius Venture Capital; Pacifica Fund; Summit Accelerator Fund; Sun Microsystems Venture & Strategic Investments; and Bessemer Venture Partners. The company’s probably been cash-flow positive (or near to it) since late 2005.

It’s tempting to view this acquisition as an admission that fighting spam is a problem requiring dedicated resources.. the upshot would be that GoOgle will certainly be able to increase the efficacy of their internal spam-fighting strategies. But that ain’t the whole story. Google likely does have the expertise to effectively eradicate spam flowing into their G:Mail network, but of course it could also benefit from more advanced strategies developed by a more evolved spam fighter, for sure. But that’s not worth $625M.

More ominously for Microsofties, Postini’s customer list (which likely numbers into the tens of millions of mailboxes) makes an excellent channel-to-market for a business-class managed email service as G:Mail moves into that particular high-value market segment. While Postini’s mailbox numbers are inflated because of some of their early success with ISPs (FrontBridge was totally focused on Enterprise), each of the “big four” managed email security vendors did a fairly good job at optimizing their sales and customer acquisition processes for dealing with business customers. Google has no experience with this at all.

But wait, there’s more: most people make the mistake of looking at the “big four” (Postini, FrontBridge, MXLogic, MessageLabs) as purely anti-spam/anti-virus companies. And fair enough: that was the crux of their value proposition as the spam problem flowered out of the late nineties and hit home hard starting around 2001.

But the big undercurrent in the managed email services marketplace after the anti-spam value proposition started to blow over in 2005 actually became compliance archiving. Big companies were spurred by the penalties imposed on company directors and board members by Sarbanes-Oxley combined with the NASD 3010 and SEC 17a requirements for public companies. Also, an increasing number of court cases have relied on email discovery to substantiate their cases, which if your archives are on CDs or Tapes can be very, very costly. That’s why FrontBridge purchased MessageRite in 2004, and it’s why immediately thereafter their competitors set forth beefing up their email archiving and retrieval stories. Google hints at this in their acquisition FAQ:

” .. businesses have been forced to choose between innovation on one hand, and these backoffice mandates on the other. In effect, many businesses use legacy systems not because they are the best for their users, but because they are able to support complex business rules.”

They made an even stronger statement to this effect on the conference call. As El Reg reported:

“In a conference call, Girouard added without giving figures that there had been “significant interest” from large business about Google Apps but that complex security requirements and meeting regulatory compliance have acted as a barrier. “Up to date we have focused on partnerships to resolve these issue but we recognised it’s in our own interest to pull things together” by buying a technology that can boost its hosted apps business.”

The archiving and electronic discovery driver in particular plays well into GoOG’s current capabilities — massively-scaled indexing, searching, and storage of data — particularly well, don’t ya think?