Wednesday, September 29, 2010

Neo-proprietary tactic considered harmful to open source


The Freemium model is a marketing model used by lots of service based companies. The principle is to offer a product "for free" supporting it with or without advertising and offer value-added services to "premium" customer that pay a fee. Very popular amongst startups and recent technology companies : skype, linkedin, rememberthemilk, etc.

At some point, one could even argue that the most successful open source company (RedHat) is very closed to this model : they offer a great product for free (the RedHat Linux Distribution) and monetize services of only a small percent of their users.

Fauxpensource has several definitions and even if this is not yet a widely used term. Some synonims are open-core or neo-proprietary. Neo-proprietary is the term I will use in this post as there is no common sounds or part with Open Source.

Somehow, what happens when you marry the two concepts together ? Well, that's somehow the subject of this post.

Open Source various dimensions for neo-proprietary +freemium companies 

Open-Source has several dimensions not only a economical/marketing : a healthy project has all those aspects covered/fulfilled : technical, political, philosophical, economical, social, ethical. My point is that, more and more, companies are using open source only for its economical value : as a cheap and fast way to build a community of users (free) and, more importantly, customers (paying) that allow them to replicate the business model of traditional software editor.

In some respect, the market is expecting this exact behavior : entrepreneur are now advised by the investors/VC/lawyers to make sure that everything is covered from contributor agreement to free and paid users agreement.

The (not so) hidden plan is simply to have an exit strategy based on an acquisition so that any number of closed source/traditional company can integrate/use the intellectual property (lawyer language here!).

Today, being open-source is seen as the fastest track to build a freemium (aka potential customer) and premium (aka real customer) "community" (much more sexy than customer base).

I will argue here that this is not really open source and it hurts the open source brand as a whole : those startups (and there is a lot out there!) do no take the time to develop their :
  • ethics. For contributor : yeah yeah, open source is very important for us : please sign this contributor agreement that allow us to re-license all your contributions as we see fit forever, etc.
  • political. Most of the time, there is no political structure, only a business structure that dictate the politic based on short term goals or based on the exit strategy (the exit strategy is not communicated to the employee or customers but all the structure is designed to maximize the value of the product)
  • technical. Well, most of the time open-source is selected for technical excellence. In those companies, an open source technology is selected based on license flexibility/compatibility, not based on technical excellence.
  • social. For products, the goal is to have paying users, not to build a community of like minded people. Same thing for employee, developers/contributors. The social aspect is simply a mean to maximize the profit when the company will be sold : it is a honeypot to attract paying customers.
  • philosophical : well, long term goals do not match very well with VC with an agressive exit strategy.

Of course, the economical (and legal!) aspects are very well developed : this is the main (only?) concern of the shareholders (entrepreneur, management, VC, etc.) : it is the only measured indicator (burn rate, market value, conversion rate from free to premium customer, etc.).

Why is this harmful for Open Source?

Well, let's follow a customer experience in this context. He knows a bit about open-source and find the model interesting. He like the idea to have a company that can provide value with a service offering or an add-on offering.

He tried the open-source/freemium edition and later on he decid to purchase the whole package. Then the company is acquired by Orache Corp. and everything is turned upside-down : the "product" can become proprietary again, price can be doubled (or worse) or the service agreement terms completely changed.

For this customer and for the market, there is no difference between this and any proprietary software : the community is, most of the time destroyed by the acquisition, the business ecosystem as well, the long-term strategy of the project can be completely changed and nobody has a saying about this : everything has been carefully planned by the buyer.

The customer has been deceived : he trusted a brand (Open Source) but has been manipulated into something else (in this example : becoming customer of a proprietary software). There is no difference between this story and the numerous stories involving dissatisfied customers that are hostages of the numerous mergers and acquisitions that occurs in the IT industry since its inception.

In fact, the mistake made by those companies is to restrain open-source to a technical, economical and legal questions only and not to consider the complete scope of Open Source namely the ethical, political, social and question.

I learned today a the open-source think tank in Paris that 90% of companies get acquired : this is the most frequent exit strategy for most of the companies and there is nothing wrong with this. Being "Open Source" should be more than an exit strategy for a start-up and that labeling those companies/projects "Open Source" is misleading for all the open source community.
 How can we, as the Open Source community make sure that  our brand still carries feelgood and positive values ?

I think that labelling properly the "neo-proprietary" companies would be very helpful : even if they respect the open source license, they clearly don't intend to respect the spirit of open source. In order to do so, proper criteria accepted by the community have to be defined and the term has to be used extensively by the community.


  1. I see people refer to what you call "Neo-proprietary" as "Open core" a lot these days. As in, there's a core that's free software, but to actually use it in a real business environment you need to buy a significant amount of non-free components as well.

    I like the Neo-proprietary term though, it describes better what it actually is!

  2. If an open source small business sells itself to a larger business and in the process becomes proprietary then a FOSS foundation should logically be setup as a parcel of the seller. When something FOSS is sold off an then a new asset is received in return to the foundation. Should that be included in a new GPL license?

  3. @Jonathan : Yes, "open core" is used as well. I think we should never employ this term as it is a part of "Open Source". You can not be "half open" ;-)

    As a consequence, the "open core" term contribute to the brand dilution. So I really think that we should stick to the "neo-proprietary"

  4. @Anonymous : Great idea ! I really think the FSF or Linux Foundation should do this : scan the web and host all the published code ... forever.

    However, if the software is not functional because the freemium version is not complete, the source code is not that useful and you will need additional development to
    have a useful open source software.

  5. Peut-on classer Google parmi les neo-proprietary? Personnellement, je le crois.

  6. @André Cotte : Je ne pense pas. Google a développé toute une série d'applications propriétaires en utilisant des outils open-source. Google ne s'affiche pas comme une compagnie "Open Source".

    Pour certaines technologies google a créé des projets qui sont open-source et contribue à de nombreuses technologies open source...

  7. Benoit—not only open source tools, but also heavily relying on open source code to which Google has applied substantial modifications and enhancements, none of which get shared back with the community because they're never actually "distributed" under the terms of the GPL...

  8. What about forking ? If a project (company) gets aquired the community could always fork the GPL code and create a community driven version ...

  9. @David "Lefty" Schlesinger : I think, we all would like to have google release 100% of their code. I don't think it is going to happen any time soon ;-)

    They completely respect the letter of the software they use. We can be sad about it or write more Affero GPL V3 code.

    My point here is that they respect the license of the software, are strong supporter of the FLOSS movement and and contribute back lots of improvements to lots of FLOSS : they don't pretend to be an open source company. As such, they don't abuse the "open source" or "open" term.

  10. @robpelu : Well, for a software that has clear value to the users, it will be the case. But the "neo-proprietary" or "open-core" is a whole different story for me : in some case, we can consider that the core does not provide any useful function. This is, sometimes, a nice piece of code and of engineering but that's about it : it provides no value to their user.

    Also, because of this, contributing code that create value is, most of the time, difficult as it goes against the business model of the company.

    Yes, you can still fork, but you will only have a nice piece of software without that much (if any!) useful function for its users : it is difficult to create a community in this context.

    For sure, it takes much more effort and energy than to fork an open source project that includes open-core and open-plugins and open-documentation, etc.

    All in all, the chances of success of the fork are much lower than for a complete open-source project : indeed, that is exactly the goal of the "open-core" like companies : to create barriers (exit barrier in this case) and be able to re-licenses everything in order to sell it to some larger buyer...