A fascinating debate is starting to unravel about SaaS and the Opensource movement. (Hat off to Ben Kepes for starting it, as a fellow Kiwi I love seeing this kind of leadership). I really like the analysis and framework used by Sinclair Schuller. I totally agree with his statement that SaaS and Opensource are not mutually exclusive. In fact I think that they are natural bedfellows.
If you buy into Will Price’s analysis, SaaS companies could fundamentally benefit from the support and passion of Opensource development. If you could reduce the time frame to liquidity from the 1.6x inhouse timeframes by having more developers involved, if you could reduce the amount of capital required by using the Opensource community surely this would be a great thing?
Bob at Smoothspan mentions some questions that would need to be worked out like how do you stop competitors using your code and multi-tenancy / multi version issues.
Bob’s points are valid, but to my mind they represent a dichotomy that exists in the SaaS / web 2.0 world.
“How can operate in a collaborative open world and still make money”
SaaS purports itself to be new world, but it still retains some old world traits.
- The "secret source is in the code",
- customer lock in is important",
- you can’t have unlimited choice.
And frankly all of these are solid business reasons.
But couldn’t it be that the secret source is the customers you have, their satisfaction levels and happiness and willingness to recommend the SaaS to other customers. A critical part of this happiness is them having the ability to take the code and make modifications to it. To my mind one of the benefits of SaaS is ease of customisation.
Not being funny, but if you could get your ERP, CRM or even email customised and running perfectly for your business in a SaaS model using this kind of customisation and the usual plethora of widgets and modules the Opensource community has, why would you want to upgrade? You have the flexibility to change it with your business.
To my mind the key thing for this to happen though is a service delivery platform.
This platform would server as an aggregation point for the Opensource and wider ISV community.
The benefits of this approach as I see it are:
- You, as the SaaS company get the speed of development and liquidity benefits mentioned above
- You, as the SaaS company also get a bunch of smart motivated people who are out there creating the perfect solution for specific verticals (multi-niching), thus making if fit for a much wider group of customers.
- In some instances this could be the customer submitting their own version of the SaaS application optimised for their business. How powerful would that be? You could even attract SaaS customers into the ecosystem with the promise that if they do submit ‘their version’ they make some money on it.
- You as the SaaS company get customers promoting the product on your behalf. If you follow the BT Tradespace model to its natural conclusion, you will have people within industry on a forum talking about your product! And this is really powerful as in my experience SMB’s normally don’t know what questions to ask, are under-informed about what the hooks are and completely overwhelmed by the amount of choice available.
- The other benefits of forums, especially relevant for SMB’s is that they don’t have an IT department, and the way forums run down bugs or ‘how too’ is incredibly important to that segment.
- If you did the rating thing for modifications or extensions in a Opensource way (check out Joomla.org), it helps when deciding on which to install.
Force.com (Apex) is close to delivering this type of functionality, the missing bit is the open standards as far as I can tell.
This is all fairly new, I’m sure I’m missing some things out her but I can’t see too much downside
UPDATE: I thought about this more over lunch. If what i say above is true, the downside would be that the current prevailing SaaS model is wrong. The model above is more akin to software virtualisation in its delivery than the one big customised software platform… or am i wrong?