What lies beyond the open core debate?

The current debate about the open core licensing strategy is as tedious as it is predictable. As soon as Jorg Janke published his excellent post on the lessons learned from the “failure” of Compiere’s open source strategy it was inevitable that someone would argue that the reason behind the company’s failure was the open core strategy itself.

Simon Phipps rose to the challenge, in doing so also countering some recent statements in favor of the open core approach from Marten Mickos. While Simon makes some very valid points, and Compiere’s strategy was undeniably open core, it does not necessarily follow that all open core strategies are doomed to fail (as Jorg himself stated “execution is everything”).

Besides the debate about open core, as Stephen O’Grady rightly points out, is futile. We have been over the debates about freedom many times before but ultimately, as I recently discussed, it will be market forces that decide whether open core remains a winning strategy in the long-term.

However, the whole debate on open core is trapped in an assumption that proprietary products or support/subscription services are the only ways to generate revenue from open source software. I am much more interested in emerging business strategies that offer alternatives to traditional product-led revenue generation and explore the potential of new software usage models.

An example is the strategy being followed at Appcelerator. The company is best known for its Titanium open source application development platform for creating mobile, tablet and desktop applications.

Titanium is freely available under the Apache 2 license, but the company also offers two paid subscription editions, Professional and Enterprise. Rather than adding proprietary features, Professional and Enterprise add early access to new features, premium support and training (and in the latter case critical issue resolution).

They also provide access to more data from Appcelerator’s Titanium Analytics – an analytics platform that provides application developers with data to track and measure adoption, as well as session data.

As well as developing Titanium, Appcelerator has invested in creating its own analytics engine that is capable of delivering the sort of session and usage data you expect from Web applications on mobile, tablet and desktop applications.

Users of the free Community Edition get one week of data free, while Professional Edition users get six months’ of data as part of their subscription. Enterprise Edition users get 36 months’ of data.

What is most attractive about this strategy is that it manages to provide additional value to paying subscribers without actually witholding any of the features or functionality of the core product from the community edition users.

It is also evident that the concept could be applied in other areas – while Titanium Analytics has been developed specifically to deliver user and session data based on applications developed by the user, there would appear to be similar opportunities for open source software providers to deliver users with analytics data related to software performance, usage and other key metrics.

Open source software monitoring and management services delivered as part of a subscription package are not uncommon. Could it be that the key to differentiation in the long-term is not products or services, but data? That’s one possibility. What are the others?

Tags: , , , , , , , , , , , , , , ,


#1 Javi Vázquez on 06.30.10 at 6:47 pm


I appreciate your approach to services, associated to open source products, putting the focus on the data.

However, I wonder if the question “Could it be that the key to differentiation in the long-term is not products or services, but data?” is not going too far, trying to deal with data apart from products or services:

Data by themselves are, most of the times, useless for the clients (e.g.: server logs). Therefore, a provider needs to gather, process and present the data to its clients in a way that is useful and worthy to pay for taking a look into.

For example, we (disclaimer: I am COO at eBox Technologies) offer a “Virtual CIO” package (reports including use and performance of network services, alerts, etc.) bundled with the eBox Professional Subscription. Is “Virtual CIO” an example of data as you consider it? Or is it just a complementary service adding value to an open source product? It might become a discussion on semantics…

All in all, I completely agree with you on not considering the open core approach (community vs enterprise edition of an, at some extent, open source product) as the _only_ way to generate revenue from open source software.

#2 Matthew Aslett on 06.30.10 at 6:56 pm

Hi Javi,

Thanks for the note. Yes, to be clear, the data would need to be delivered as part of a subscription package – what I foresee is different amounts/types of data being the differentiator between subscription options, as in the Appcelerator example.

#3 Giovanni Tirloni » Blog Archive » The Open Core Debate on 06.30.10 at 7:34 pm

[…] Core is Bad For You Open-Core: The Emperor’s New Clothes While lies beyond the open core debate? Open Core is the New Dual License The Road to Closed Source Software, Eucalyptus Marten Mickos says […]

#4 Jack Repenning on 06.30.10 at 9:13 pm

Arguably, Titanium is not Appcelerator’s “core” at all, but a “complement” (specifically, an “infrastructure”) without which their real business could not exist, or would be far less successful.

And Appcelerator Titanium has hit what looks like an ideal market division for open-infrastructure business:

The open part is a developer tool, so opening it to its natural customers has a high chance of returning actual value (code and patches) to Appcelerator (where opening the code of other products can sometimes prompt developer yawns). But actual money flow for developer tools has proven over the years to be a very chancy, spotty affair.

The closed part, on the other hand, is a business / marketing / executive / strategic tool: customers who are less likely to provide helpful patches, yet more likely to provide cash flow.

So it makes sense for Appcelerator to open the part they do, and hold closer the other parts they do.

#5 451 CAOS Theory » Open core is not a crime on 07.02.10 at 10:44 am

[…] core is not a crime Matthew Aslett, July 2, 2010 @ 10:44 am ET One of the reasons I described the current debate about open core as futile is that there seems to be no hope of it ever reaching […]

#6 451 CAOS Theory » Do customers want open core? on 07.08.10 at 1:54 pm

[…] customers want open core? Jay Lyman, July 8, 2010 @ 1:54 pm ET There is renewed and meaningful discussion going about open core with several good insights and arguments: Simon Phipps, Daniel Radcliffe, […]

#7 The Open Core Debate | Giovanni Tirloni on 08.31.11 at 7:53 pm

[…] Core is Bad For You Open-Core: The Emperor’s New Clothes While lies beyond the open core debate? Open Core is the New Dual License The Road to Closed Source Software, Eucalyptus Marten Mickos says […]