Monday, June 25th, 2007

Meme Games: 3×3 Lessons for Open Source Firms


I was tagged by Stephen O’Grady… while I generally avoid replying to such meme’s, I then saw Luis replied and that put the rest of us tagged on the spot ;-) I am also replying because I probably see a different side of the equation that could potentially help others understand how the IBM-like vendors look at open source firms. Hopefully there may be at least a small level of insight that someone can benefit from in this response. I’ll put the usual disclaimers around this – take it “as is” with no express or implied warranties for fitness for any particular purpose (or merchantability).

I will preface that this meme is about “open source firms” which I assume means companies/profit seekers leveraging open source software as a model to grow – not those who take on the humble task of building free software in a non-commercial intent model (i.e. Apache, Eclipse). Having replied to this meme, I get to name the next round of victims. For that I will now tag Canonical’s Mark Shuttleworth, kernel hacker Gerrit Huizenga who participates uniquely with other Linux firms in the community, and finally sogrady’s partner in crime over at Redmonk, Coté. So without further adieu, here’s my quick/dirty 3×3.

3 “Do”s:

1) Do Bring Crisp Customer Value to the Table: Define a value proposition beyond simply relying on being “open source” – there has to be some reason your value proposition is better (i.e. Firms in your industry constantly struggle to integrate X and Y while customizing for individual business units. Our open, modular architecture can be adapted to meet your individual business unit CRM requirements and integrate with X without expensive, lengthy customizations of the entire app) Hitting on a pain point of “closed” alternatives works well too – ask a PBX user how much they love their supplier(s)… turn the pain points (licensing, payment models, etc) into your value prop

2) Do Make Friends, Not Enemies: I once had a _smart*_ once tell me in an opening introductory sentence at an O’Reilly event his job was to use open source to wipe out millions of dollars of IBM’s business until IBM realized they’d have to buy him. That was his first sentence to me ever! – not “Hi, I’m XYZ from ABC, I’d like to find ways to partner with IBM”. The IT business is driven by partnerships, friendships, loyalties, acquaintances, and networking. If you are into open source to make a profit, partner like mad – if you position yourself as the enemy, expect to be killed or worse, ignored and slowly put out to pasture. The person who approached me in that manner is now struggling with the latter issue… I suspect his initial investors won’t hang around long as he still tries to make himself relevant.

3) Do Make the Right Friends – Partner with the Right Partners: The open source firms with a strong management team “get” partnering. They partner with other firms that can enable business and community growth. Just because someone like IBM invests millions in open source communities, Linux, and the community does not mean every open source firm should bet the farm on partnering with IBM. Yes, partnering with IBM has advantages, but IBM can also suck up your precious resources simply navigating the huge array of virtual teams. IBM may also have a differing strategy (and every strategy out of IBM is not necessarily the one that wins in market – you can be different). As an open source firm, your goal is to find the right partners who are naturally aligned for partnering with you. Don’t accept partnerships that have no investment on one side – they’re doomed. Both sides must be aligned and invested. How will your partnership not only create more business for you, but also generate new revenue or reach a new customer base for your partners. I’ve had this discussion with the head a particular open source firm on a few occasions – partnering is by definition mutual – how do we both benefit from partnering? (otherwise, I can’t justify investment and maybe I’m not the right partner for you) Another IBM partner with an open source product just closed another round of investment for ~$25M – they know how to partner in a mutually beneficial model. There will be a lot of news coming in this space…

3 “Do not”s:

1) Do Not Shy Away from Being a Commercial Business – too often companies pitch how “open” they are. While that’s great, how do you make money? It’s as if some open source firms are afraid to discuss how they make money off of free software (free as in speech). This is usually a good way for vendors to weed out the bad firms – if the firm is shy about discussing this, they probably have a weak value prop, and there’s probably more of a “hidden hook” to make customers pay than offer something customers want to pay for. IBM has for a long time been investing substantially in Linux, open standards/source based software, and open source communities – I can’t recall an executive who did not clearly, up-front, and simultaneously articulate how IBM generates revenue from any of those ventures. If you set the tone up front, no one will be surprised or upset later. And if you have a strong value prop – see point 1) – customers will expect your valuable product/service comes at a reasonable price. And one last point is “free as in beer” only gets you one round – downloads do not magically turn into profit later. I don’t care if you have 8M downloads – while that’s great for showing loose interest in your technology, how many are paying you? How many have built a dependency in their applications on your technology? How are you going to generate the revenue needed to fund the next version of your technology?

2) [If you are trying to be the standard as in platform], Do Not Try to Control/Dominate/Dictate an “open” Community Project. I’m not talking about MySQL here – they’re an application/database component. Rather, think about platforms like operating systems, dynamic server languages, portable runtimes, etc. that customers use as platforms to build on. There are many of these succeeding today (RHT, NOVL, PHP, Apache, Eclipse to name just a few). There are others that may have an open source license, but the firm tries to dominate with restrictions and provisions in contributor agreements, including hidden hooks in license terms, or by bundling in non-free (as in speech) components. These firms typically then claim to be leaders and their platform is growing in adoption … yet for some reason… no one partners with them – see point 2) – and they carry the entire development expense with no community-scaling benefits. They invest the R&D and in the end, their “community platform” is just another vendor product that rings well with the vendors existing customers. Consider Eclipse – while it may have started with IBM dominating the “participant list”, it takes a serious approach to vendor-neutrality and open participation for BEA, Oracle, Sybase, Zend, Actuate, Compuware, SAP, CA, and Borland to all join IBM on the Strategic vendors list. The same applies to Linux – look at the Linux Foundation membership list.

3) Do Not Ignore Intellectual Property; Do Not Let Intellectual Property Stifle Your Innovation – First, you cannot ignore the effects and implications of intellectual property decisions. First, there’s the license – not all software must be GPL, but there are serious, practical issues if you decide to create the “XYZ Corp Open Source License v1.0″. License appropriately for the type of developer and user community you’re trying to foster. Second, patents – take a political stance, but don’t get caught staring at the clouds waiting for Congress. File for patents if you can, what you do with them after is up to you, but the Patent Commons, OIN, or some other entity is a great place to house them. Also, don’t be stupid and violate an obviously valid patent. If you’re starting an open source project in a technology area that has established players – hire an attorney to guide you through the minefield. It can be done. Third, figure out the trademark situation. Badgeware is one issue, but not creating a trademarked identity that your “open community” can use in viral marketing and “I’m proud to be an XYZ-user” situations is sheer nonsense. You need to establish legal trademarks, then set clear guidelines for trademark usage up front, make it open, and don’t use your trademark to make people pay you – IMO it’s a poor business practice and creates unnecessary frustration later on. And finally, don’t let IP hold you up. Everyone knows this field is wrought with nonsense and arcane approaches to IP. If you can create value, just hire a reasonable IP attorney to walk you through the field. Take some basic steps to be safe, but don’t spin wheels and tie up valuable resources trying to analyze (or let an attorney make you analyze) every conceivable, possible, extreme angle. IP generally comprises of Patents, Copyrights, Trade Secrets, and Trademarks – use them wisely.

If I had a 4×4 meme, I’d also cover open standards and the importance to lead in driving these open standards as part of your model (see what happened with Spring for instance).


One Response to “Meme Games: 3×3 Lessons for Open Source Firms”

  1. June 26th, 2007 at 1:26 am

    tecosystems » links for 2007-06-26 said:

    [...] Michael Dolan Dot Com » Meme Games: 3×3 Lessons for Open Source Firms excellent response from Michael on the open source business meme – thx for playing, sir (tags: MichaelDolan opensource business economics meme) [...]



Please leave a Comment