Blog post

Open Source Is Trending Towards Customer Obscurity

By Brian Prentice | December 09, 2009 | 25 Comments

Let me ask you a question. I’m going to assume that at some point over the last twenty years you bought a car. So, how important was the car maker’s use of just-in-time manufacturing to your purchase decision? I’m going to go out on a limb here and say it was of no consideration at all.

Well, I think we’re fast approaching the point where open source to software will be like JIT to automotive manufacturing. While it will critical to the producers of software, woven into the fabric of its operations, it will be of no importance at the point of consumption.

As hard as this might be to accept, open source is not a value proposition in its own right. That’s because the market for software is driven by four questions customers ask:

  1. does the software do what I need it to do (feature bake-off)?
  2. who will help me when I have a problem with the software?
  3. who will maintain the software and how much will that cost?
  4. will I end up being beholden to the software supplier?

Open source is only relevant to the extent that it creates meaningful differences in these categories. But ultimately, customers only care about the differences – not the way it’s been achieved.

This point was brought home to me again in a recent conversation I had with Cheryl McKinnon, Chief Marketing Officer at Nuxeo. As she highlighted to me, Nuxeo endeavours to frame themselves as a viable Enterprise Content Management provider that happens to utilize open source software – not the provider of an open source Enterprise Content Management solution. This marketing approach is not unique to Nuxeo and is based on a pragmatic realization of how customers acquire software. No amount of open source spin is going to compensate for a substandard solution. The product still has to be functionally competitive within its technology category.

And while open source can have a positive impact on support, maintenance and vendor lock-in, those effects are not universally consistent. Different open source projects have different dynamics based on its diversity of code contributors, licensing provisions, and the extent to which support is required and the willingness and availability of a community to do so (just to name a few). These dynamics, in turn, can be effected by the organization which is using the project for their commercial solution. So again, astute customers will not just assume that the existence of an open source license creates better support, maintenance and lock-in conditions. They will drive each individual provider to prove it.

Open source might have a feel-good glow today. But the more enterprises consider open source software the more that glow will fade under serious scrutiny. In that environment, generic open source value statements ultimately provide little meaningful benefit to a software provider. Or worse – they come across as patronizing, mom & apple pie statements. The more that’s recognized the more open source will be banished as a selling proposition by marketers across the industry.

In years to come, open source will be prevalent across the software vendor landscape. But they’ll be selling the sizzle, not the sausage.

Comments are closed


  • Simon says:

    Love the JIT metaphor . In it’s day, with Just-in-time manufacturing was cutting edge and now I can’t imagine there are any sensible manufacturing companies that don’t use this process.

  • I think the claim of those advocating OS is that your point 4. will have an absolute “No” answer if you use OS. That’s of course debatable but, if you chose the OS product wisely, I think it’s generally the case.

    The same advocates tend to also tend to answer all of the questions with “You can do it yourself”. That’s of course just wrong.

  • Great points, Brian, and I wholeheartedly agree. I hope these considerations will soon make their way through those legions who keep pushing for governments to give priority to open source.

  • Sindri – I agree. I think that the ability to mitigate lock-in is increasingly the core value proposition of open source. But open source is not without its own lock-in risks. For example, does a single provider dominating available support resources count as lock-in? What happens if there is no one willing to fork a project that is being abused by a key provider? We’ll need to learn how to identify and manage these scenarios as OSS continues to mature.

    Thanks though – great comments!

  • Hi Simon – and I think in another couple years we’ll be saying the same thing about open source – no sensible software company will not be using the process.

  • Mark Walker says:

    Interesting post. I notice that the majority of people see Open Source as free beer NOT free speech, so I think the cost on the box is a bigger draw than your piece suggests.
    BUT I also see people choosing software for all kinds of ill-conceived reasons and living with the consequences of choosing the wrong solution and/or the wrong supplier and/or/usually both.
    Choosing Open Source is not a guarantee that you have made a good choice, but good decision-making may lead you to Open Source.

  • Arunkumar N.T. says:

    Customer obscurity? I doubt it; on the contrary many customers are getting more inquisitive about the “ingredients” of the product / service they are paying for – primarily because analogous to JIT in manufacturing (which promised, in some ways, cost-cut passing on to the customer through smart inventory mngt), adopting OSS frameworks imply (arguably though) faster assembly of the solution at lower costs of development. Like anything disruptive, OSS will very soon evolve to address issues of “ownership” and maintanability, as a response to business models undergoing massive changes – will ‘software’ continue to be looked upon as an asset? Or as a consumable service that can be switched over to the least-cost, most-current-in-features, flexible service provider reflecting business/market trends?

  • Rudy Godoy says:

    I support your point. I’ve seen different tech companies in the FLOSS landscape or ecosystem trying to differentiate among their market selling the technology. The big mistake is that their customers do not buy the technology, they buy the final product (which, again, happens to be build on a open-source software stack).

  • Craig Tobey says:

    Open source has the benefits of the network effect. When you connect your proprietary software to open source, your software instantly benefits from the entire network of people around the open source product. In a way, you are also signaling that you are lower cost as you are using “community” resources. It is a disruptive technology.

    Here’s another analogy: do you care if your automobile vendor uses robots? Nobody really does – they all use them now. So, does that mean it is not important? Not at all….anyone who does not use them is crazy. And, in fact, companies used to show how they used robots when it became big.

    Open source does provide meaningful differences – those differences are the reason why your software is better…and why people will adopt it.

  • Brian,
    A pragmatic post and a good comment about vendor lock-in, however as others have pointed out this can be avoided by careful product selection. There was a blog post recently (although I can’t find it offhand) comparing OSS with the organic food movement and how many companies are ‘bending’ the ‘organic’ label by doing the very minimum and complying only as far as they need to to be able to use the term for marketing, and not embracing the full spirit and culture.

    This results in a dilution of the meaning of organic. Same with Open Source, and the blog post suggested a self certification criteria to score particular elements of ‘open sourceness’. One key element was to look at the governance and control of a project. Is it all under control by one company (the risk if which you highlight above) or is it controlled by a community or foundation.

    In the CMS world for instance you have the likes of Alfresco or Ez Systems who whilst are open source by license are still controlled, governed and supported by one company. In contrast you have the community open source projects such as Plone or Drupal who are under diverse control. In the case of Plone by the Plone Foundation, which provides a central owner of the IP, but which is governed by an elected board of directors from the community.

    So as you say, maybe a buyer will not be looking for the words ‘open source’ on their next product selection, but they should (especially in current financial climate) be looking for diverse control to lower risk of lock in or vendor failure or takeover. Open source is the only way to provide this assurance transparently.


  • Andrew Back says:

    Some good points here but I’m not sure I entirely agree. Sure, the end result is what’s important, but our sensitivity to lock-in is becoming increasingly heightened and our freedom a growing concern. Meanwhile many organisations are starting to realise they can use F/OSS without support contracts in place and the sky does not come down – they’re doing just fine and can dispense with the costly Vendor’s New Clothes. Of course, this is not universally true and there are a great many factors to take into consideration before going it alone.

    And what about home grown F/OSS? Many enterprises are open sourcing code of their own creation and fostering collaboration that crosses the company firewall. Simultaneously delivering on behalf of internal customers and a wider community, and benefiting from anything from peer review and bug reports to reduced development overhead. Here open source technology, principles and delivery model are at the centre. Vendors are inconsequential. Sure, they can use the code, contribute to it, sell it, whatever. But they are now a peer and nothing more. It’s no longer a consumer/creator world.

    @Andrea Di Maio We’ll have to agree to disagree. My personal view is that public money should fund the development of software that is in the public interest, and I subscribe to Dan Bricklin’s assertion that “the structure and culture of a typical prepackaged software company is not attuned to the long-term needs of society for software that is part of its infrastructure” (see his piece Software That Lasts 200 Years). So, even where F/OSS solutions are behind their proprietary counterparts I’d rather we took a long term view and invested in building ‘societal infrastructure software’, than taking the vendor solution because it’s easy and available now.

    @Matt Hamilton I like the organic food comparison. Thankfully, once again people are starting to get wise and to realise when the open source label is little more than lip service, and to also question project governance. Sun learnt the hard way with OpenSolaris and its associated community… At an event in London earlier this month Bruce Perens suggested that there had been “no significant community development of core product during the tenure of MySQL AG”, and you might wonder if this is down to it clearing being their party and the dual licensing strategy requiring reassignment of copyright on contributions etc. So, whilst you can fork MySQL (and I’m more on the side of Moglen than Mueller in this recent debate over the power of the GPL in light of the Oracle acquisition) that is has little community with experience of developing core is a concern. Interesting to observe how the more democratic Postgres is growing in popularity (doubtless for many other reasons than implied here)…

  • Ken Wasetis says:

    Great discussion and it’s definitely what our firm has experienced in the market. There are plenty of reasons that different clients might select an open source software solution, but in our experience, the top 3 are:

    1) The product does what we need it to do (or it wouldn’t be considered in the first place)

    2) The client (with help of an integrator or their own technical staff) can customize and extend the product to twist and bend to the client’s needs and integration requirements better than can be done with a commercial tool whose source can’t be modified. This is typically a more key part of the ‘free’ (as in freedom) component of open source than the lack of license fee.

    3) It is in many cases also free, as in, there’s no license fee (but there are of course examples of ‘Enterprise’ options with warranty, documentation, etc. from the likes of RedHat that are commercial options.)

    For some clients #2 and #3 are flipped, but #1 is always the same, as the original post above points out.

    We have various examples from across the distribution spectrum in our projects:

    1) The Plone CMS – no license fee, democratically controlled intellectual property, project management, and development, but wouldn’t even be at the table if it didn’t provide the same enterprise CMS features as commercial CMS tools (easy web content publishing, workflow, versioning/rollback, fine-grained permissions managemenet, blogs, forums, etc., etc.)

    2) SugarCRM – While there is a ‘community’ edition, most clients leverage the ‘enterprise’ version or the SaaS/hosted option, for which there are corresponding fees, but it still provides a cost savings (versus Salesforce or other commercial CRM tools), and has the option of extensibility via the source code (less an option with the hosted plan, of course.) This tool is also controlled primary by one company, but has a healthy ecosystem of add-on modules (integrate with Hoovers, LinkedIn, etc.)

    3) Postgres relational database – Quite democratically controlled, very robust, and powerful. Extensible, with modules to allow scripting in your favorite OS language.

    I think these are all good examples of how the various degrees of control by one company, the level of democratic control of the project, and the degree to which clients are willing to pay for something that in theory is free (in the case of wanting support, enterprise features, or a hosted option.)

    In the end, it just confirms this original post – that whether the tool is open source or commercial isn’t the most important thing initially. But in our experience, the fact that it is open source and provides additional flexibility in the long run is the deciding factor for choosing an open source solution overall.

    It isn’t as much about ‘vendor lock-in’ as it is about wanting to avoid being painted in a corner with the solution and being held up from integrating with other applications, scaling up on additional servers without waiting for additional licenses to be purchased, as well as being able to select from one of many possible vendors who know the tool.

    Do others find the same?

  • Arunkumat N.T. – well, I think we’ll have to agree to disagree on one point. I have never seen customers caring about the ingredients. In fact, I see them more concerned about things like indemnification so they never have to worry about the ramifications of the ingredients being bad.

    But, I do agree with you on one key point. I think there is a clear trend where software is moving towards a new type of service model (distinct from traditional IT services) and away from “bought and sold” assets. Mind you, we don’t buy software today – we license it. We’re further down this service route than most would admit.

  • Matt Hamilton – thanks for the comments. The blog you’re referring to is located at

    I’m glad you pointed this article out – I might just respond via my own blog post because I think this analogy is a great example of where some people with an interest in open source are marginalizing themselves. Keep in mind one thing. Mixing organic and un-organic food – be it in the actual ingredients or in it’s production – means the whole thing becomes un-organic. But that’s not the case with open source. You can blend either the ingredients or the process with proprietary code and the open source code isn’t effected. Ultimately it comes down to what the customers really wants and is prepared to pay for.

  • Andrew Black – thanks for your comments. I appreciate you adding so comprehensively to the conversation. Two quick points on your comments:

    “Meanwhile many organisations are starting to realise they can use F/OSS without support contracts in place and the sky does not come down” – very true, but that depends on the project. Maybe that’s the case with something like Ruby but I don’t think your average IT organization is prepared to go without support on Linux.

    “And what about home grown F/OSS? Many enterprises are open sourcing code of their own creation and fostering collaboration that crosses the company firewall” – again, a good point. But that’s exactly my point. In this case open source is important at the point of production, not consumption.

  • Ken Wasetis – you said:

    “The client (with help of an integrator or their own technical staff) can customize and extend the product to twist and bend to the client’s needs and integration requirements better than can be done with a commercial tool whose source can’t be modified.”

    So, what about commercial product that can be modified? Many modern applications have very nice architectures that allow flexible design through customization and configuration capabilities. Those are generally a lot easier and less expensive than modifying source code.

  • I could not disagree more with the sentiment of your article, but appreciate the alternative opinion you formulate, except an obvious error:

    You said: \Nuxeo endeavours to frame themselves as a viable Enterprise Content Management provider that happens to utilize open source software – not the provider of an open source Enterprise Content Management solution\

    And right across from the logo on it says: \Open Source Enterprise Content Management\. Kindof contradicts what you are saying and what the CMO allegedly said. I would expect the CMO to actually know what her company’s website chooses as a tagline, but, hey – the naive me, what do I know? 🙂

  • Irakli – yes, I noticed that too. But I’m not going to crucify them for that. Ms. McKinnon wasn’t spinning anything. Instead I believe her comments to me were based on the way in which they’re interacting with their clients. If that hasn’t flowed through to every aspect of their web site and marketing material I’ll forgive them for that.

  • Brian

    30 years of selling and supporting software have taught me that innovation remains a key factor in decisions by enterprises of all sizes to use a particular solution. SAP built their company one module customer at a time, promising to build new capability and then delivering on it. Oracle, Microsoft and others all are companies that rose to the top of their market segments through great marketing and sales, and yes, innovation.

    It’s not enough to have the features today, nor simply provide maintenance for those features, organizations want to know that the features they will want tomorrow will also be there. With cars, that means you buy and sell, then buy again. With software of course, the model is dramatically different. Your comparison with JIT is therefore surprising to me.

    I’ve stated it another way here:

    Tom Erickson
    CEO Acquia

  • Thomas Erickson – nice comments. And there’s a lot I agree with. But when you say, “…organizations want to know that the features they will want tomorrow will also be there” we need to recognize that they aren’t using most of what’s in their software already. I saw an interesting statistic a few years ago from a company that had a software auditing tool. They found that the average SAP customer used only 26% of the available purchased functionality. In that type of situation you really need to ask what the value of a maintenance agreement is when a key part of its value proposition is a whole pile of new features in the next version.

    But that’s a separate topic… 🙂

  • Brian,

    as much as I appreciate your opinion and respect it, my problem is that I did not get a feel that it was factual. Sorry, but your main quote is from a person who contradicts with the tagline of her company. I did not see any data. You say “is trending”. When I hear trending – I think “data”. The only data I see in the blog post is anecdotal.

    It feels like the post is written from a “gut feel” rather than factual data.

    Thank you for raising important questions, however.

  • Irakli – thanks for our honesty. Keep in mind that the Gartner Blogger Network is all about opinion. This is not the vehicle we use to present analysis of primary data sets – that is the other research we do as part of our core business. So you’re right – our blog posts tend towards gut feel.

    Having said that I think that it is perfectly valid to present anecdotal evidence in the context of \trending.\ I’m putting that out there fully appreciating that I will get people’s comments either agreeing with or disputing my observations. But what I can assure you is that I am not flippant with my blog posts. While I may be commenting from \the gut\ those comments are borne from many conversations with clients and providers and blended with my own views and beliefs which have formed over 23 years working in this industry. It is my hope that observations will provide some degree of insight to readers.

    There is a place for both types of observations.

  • Ken Wasetis says:


    I didn’t mean to indicate that one should ‘prefer’ to modify the source of a CMS. Each decent CMS worth its salt, whether commercial or OSS, is able to provide customization APIs and interfaces (through the browser or otherwise) that enable a site admin to build custom page layouts, custom/structured content types, and custom workflows – very basic ingredients for something claiming to be a CMS.

    What I was referring to in terms of the greater extensibility of open source CMS systems were things that are either core to the CMS (such as the administrator/editor interface/functionality) or its ability to integrate with other systems and services on the web that allow the solution to be greater than the sum of all its parts (e.g. integration between Plone and Salesforce or SugarCRM is superior to what I’ve seen available with commercial CMS tools, but even these connectors themselves are open and extensible.)

    So, it’s not that one necessarily WANTS to have to dig into the code to modify the date entry widget from his/her CMS, but isn’t having the option of modifying it and improving it, so that it doesn’t limit a content editor from selecting dates prior to 1901, for example, rather than hoping and waiting for a ‘patch’ from the commercial vendor a much better place to be?

    And, as others on this thread have alluded, and plenty of Gartner experts, no doubt have indicated, software is trending toward more of a service model and the use of virtualization and/or cloud-based hosting is certainly growing.

    So, if the commercial vendors are slow to adjust pricing models and continue to have pricing based upon a per-CPU or per-CPU basis, does this not impede the business from being able to scale its web application managed by its Web CMS? I prefer being able to start up another license-free instance of Zope/Plone and add it to the cluster, perhaps on a new (virtual) server instance in my cloud, without having to get budget approvals for additional license fees.

    It’s obvious that businesses need to be more agile than ever and it’s been my experience (we work with commercial and open source solutions and integrate the two, in many cases) that open source wins in this regard.

    It isn’t just about the ability to dig into the code. It’s about being able to turn on an dime.


  • @Brian – our chat in December was helpful for us as we worked through some updates to our Connect subscription program – some great insights that helped us stay focused on simple clear packaging of our maintenance offering. My informal thoughts here:

    @Irakli – sorry I didn’t see your comments sooner. I certainly am familiar with our company tagline, and let there be no confusion – Nuxeo is committed to being open source – it is core to our identity. But we judge our success on the quality and strengths of our ECM offering. We’d rather be known as challenger to traditional ECM vendors than just an open source option. ECM is what we do. Open source is how.