The growing interest for OSS technologies, especially from mid-large enterprises, is struggling against the quite peculiar nature of Open Source shops on the market. Give or take a few large companies, the OSS business ecosystem is built upon a quite impressive number of consultants, associates, and small shops: this has to do with the coarse nature of Open Source. People (and companies) have diffferent backgrounds and are committed to different technologies, yet even though most readers of this post will have no problem understanding why Linux shop doesn’t support Velocity, from an external (yeah, you can call it dull) point of view, there is a radicated vision of Open Source as a whole. Deal with it.
If we consider companies making up the OSS market, there is an incredible amount of knowledge and skill floating around. A good portion of the shops out there, despite being little more than one man shows, are clearly able to solve most business needs and complete successfully complex projects. However (again, this is the perception of our small part of the world) this clashes with the needs expressed by large players: the big ones are catching up and realizing that they need an Open Source strategy but still there is a clear struggle in finding the right partners to help them out.
A large corporation has a lot of issues in managing business relationships with small shops: they do hire consultants, and of course they are used to talk with large firms, but a small sized company scares them away. There are multiple reasons for that, some of which associated with practical issues (think about ISO9000/Vision 2000 compliance): mostly, however, there is a clear feeling of uncertainty associated with small shops.
As a consequence an interesting trend is emerging. Markets are usually based on a pyramid of suppliers: your sandwich is sold you by a shop buying tens of loaves of bread from someone who has bought a ton of flour from someone else who stocks truckloads of wheat products. In the traditional software model companies are used to assumes that stuff is bought from VARs partnering with a wholesale providers of technology such as Microsoft, IBM or Oracle.
Open Source based projects don’t work this way (again, with a few exceptions): the final user is supposed to get code from the source and then start looking for support and integration, delving into the jungle of those scary small shops and with no real competence to understand and measure the right partner. There is a clear lack of a trusted source, where “trusted” has a wicked association with “large”: I know this is the old “who do I sue if something goes wrong” refrain, but reality shows that this is the current market situation, at least here. The quest for one-stop solution is alive and kicking here, and traditional players in the consulting business are getting ready to answer the demand.
Of course this is impossible to achieve internally, even to them: there are just too many Open Source technologies for a single company, no matter how large, to master. The emerging model, then, tends to subvert the traditional supply chain, and there is a clear trend towards the “hub” model: a somehow OSS aware traditional consultin company gathers a bunch of small, specialized, shops, looking for the technical skills they’re missing. They continue of course to provide legal backing, a trusted brand, Vision 2000 compliant processes, and all those features large customers are looking for. The actual projects, though, are mastered head to toe by their small partners or consultant networks, with the large guys formally backing the gigs knowing little or nothing about the underlying technologies and the final customer ignoring who’s really behind his project.
This is clearly a VAR model (albeit wicked), but it’s somehow weird to see it happen in the software arena, within an inverted pyramid environment, where wholesalers are vehiculating small (and unknown) suppliers. I’m starting to wonder whether it would make sense to push this model forward, making the process more explicit. This could happen by creating OSS “hub companies” whose role would be scouting viable and proven OSS small shops while providing accountability and marketing/process overhead large customers need. The key, though, would be retaining as much transparency as possible towards both the backing network of specialized small shops and the final customer: this isn’t easy since it would require a clear perception of value from the underlying participants (something that isn’t granted) and a notable amount of trust from the customers. However, the more I think about it, the more I’m convinced that there might be a way to make this really work.