Matt Asay (Alfresco) and Stephen O'Grady (Red Monk) have been having a fascinating debate that began at OSCON last week over whether we will see a billion dollar pure open source play or not, if so when, and whether that's the relevant yardstick or not. The entire discussion is well worth the read and the links are included below.
The one thing I find troubling is the focus on the billion dollar yardstick. Every new technology shift sees the creation of new companies that can't be measured using the same metrics as the historical base. It's meaningless to compare MySQL to Oracle using revenues (the MySQL user and customer bases don't pay for the database), or installed base (MySQL is easy enough to use that it can be trivially dropped into place as a groupware or departmental database).
But then we need to ask why comparing Oracle to MySQL AB is even relevant. It isn't particularly to the way MySQL AB runs its business. It isn't to MySQL's customers. The only people that seem to need the analysis is Oracle itself so it can demonstrate to customers and investors that "everything's fine."
Drucker (as a good economist) pointed out:
A business doesn't exist to maximize profits — that's a measure of success. A company exists to make a customer. To make a market.
You have to love the simplicity of that statement. In the end it is all about customers and making them wildly happy by solving a problem they value.
Stephen argues that volume is a more relevant measure of success for an open source company, and this starts to get to the customer focus through the volume of users. MySQL has a huge deployed user base and focuses on converting users to customers. This conversion rate discussion is central to Matt's thesis, but I'm still worried there's a subtle but important aspect of the customer that's being missed.
Matt gave a great OSCON presentation last week (Making Sales while making Friends: Lessons Learned from Open Source Businesses), and his blog posting sums up the key slides best. He is struggling though with the concern he's voiced many a time that the open source applications space will NEVER have the volume downloads of the open source infrastructure space. Essentially, no one will ever download Alfresco 40M times. So raising conversion rates defines his battlefield and therein lies the problem.
Alfresco rebranded itself to an Enterprise Content Management solution through the Fall and Winter of last year. Essentially, "we are all things to all people in the content management space", instead of staying with the laser sharp focus of the best document management system on the planet at a "free" price point, and as an open source solution. Presumably, they're wrestling with the problem of value definition to their customers, and feel if they position themselves as the Documentum or Interwoven killer, they'll have an easier time of it.
That's the problem. They now have to live up to that solution, and a customer's existing investment in the entire space. While they're having success with conversions (and Matt's discussion of how they convert many customers was brilliant), they're still limiting their perceived value to the few people that consider content management problems in the enterprise.
They need to live their all-things-all-people marketing message. Indeed, for those of us that attended the O'Reilly Radar Executive Briefing last week, we were horrified to watch John Cochrane (Alfresco) give what amounted to a standard marketing product pitch for his 10 minutes of fame, identified on the O'Reilly Radar as one of the hot open source companies to watch. The other seven companies talked about what the company did, why open source was fundamental to the business model, and how their communities were central to success. Mr. Cochrane told us why the Alfresco product was better than its competition with a feature list. At the end of the session, you have to wonder why they're a hot company to watch. They didn't tell us.
The MySQL user community weren't Oracle DBAs and CIOs looking for a "cheap" fix to the Oracle spend. They were non-users of databases (or people that didn't have easy departmental access to the "large corporate database resource"). And it spread like wild fire. They didn't care that it wasn't "an enterprise class database" in the beginning. It solved their problems brilliantly.
So instead of thinking of Alfresco as "an Enterprise Content Management" system, let's for a moment treat it like a quick-and-easy document sharing system. Every project, workgroup, department, and medium sized business (or school, or non-profit, or religious organization, or community center) could drop down such a system instead of all those important Microsoft Office, Wordperfect, and OpenOffice documents being stuck in people's laptops and desktop machines, which we all know is becoming a problem. I could trivially drop an Alfresco server on Windows or Linux. Just like all those non-database savvy users dropped MySQL database servers all over the place. Millions of them. For free.
Now there rapidly comes a point, especially when multiple people in large organizations start dropping document shares down like candy, when the corporate infrastructure people get involved to do the responsible thing, and need to support and maintain the resource. And that's when they start calling Alfresco. But the joy of this is that it is likely a greater number of enterprises that will have this problem than the fewer number that thought they wanted to tackle an Enterprise Content Management problem. It's all about the area under the curve. Conversion rates are important, but I want thousands of customers calling me to convert on the hundreds of document sharing servers they're already running, not hundreds of customers needing to call me as they try to wrestle the behemoth of the ECM to the ground.
This is the ground that Microsoft already wants to go after with Sharepoint Server 2007. Indeed, they're boldly telling customers that this will be the value "lock" for the next generation. Of course, they're telling customers that are tired of lock, that want choice, that are beginning to investigate open source solutions to regain a sense of choice, and that are exploring OpenOffice on Windows as a way to manage the Microsoft spend. Alfresco doesn't need to "compete" with Microsoft directly to succeed against Sharepoint. They're already ahead of the curve here.
But it will require a sense of focus and indeed humility. It requires focusing on customers' problems (the reason you're in business), rather than investors measuring profits and growth towards a billion dollar company. Can they step away from being ALL THINGS TO ALL PEOPLE to merely being the world's greatest little document management system ... at least until the world is breaking down their door. Who needs to be a billion dollar company, when you can be a wildly successful and growing company that happens to be profitable.
The posts that began the discussion:
- Stephen's initial discussion: Billion Dollar Open Source Businesses
- Matt's response: Can Open Source Crack the "B" Barrier
- Stephen's deeper dig at the numbers: Breaking the Billion Dollar Barrier
- Matt extends the discussion with comments from Marten Mickos (MySQL): More on the Billions Question (Marten Mickos)
It seems to me that John Cochrane (Alfresco) is not trying to be all things to all people as you state. He is focusing on the "Enterprise" with his product.
He understands this space from a marketing perspective; the problems that users are facing, the key industry influencers, the competitors, etc. I believe that his intent from the very beginning was to bring Alfresco into this space, compete, and then eventually dominate the industry.
Open Source for this purpose is an enabler for getting his product developed, etc, but is not a "means to an end" as you would like to see.
Obviously John's target market and approach is considerably different than the one you would choose. It's his choice and he obviously thinks we can compete at that level.
Time will tell.
Posted by: Ken Mulcahy | 05 August 2006 at 06:29