Oracle, Google and Aras’ 226% Growth

Oleg, I suspect most engineers reading a 226% sales growth claim, without any qualifying base, would LOL. Is that in revenue or in unit sales? Is that based on previous period sales of $100K or $100M? (Intuitively, you have to assume it’s much closer to the former than the latter.) There’s no way to validate that 226% (or 26% or 326% or -226%) is accurate. Unverifiable marketing puff, quite useless for determining the success of Aras’ shareware model vis-a-vis real open source or traditional software licensing.

Oleg, you’ve often speculated how open source software may challenge traditional PLM suppliers, and in this case offered the Aras press release as support. However, this Aras press release can’t provide any insight into your question because (1) the 226% figure is useless for determining sales success, and (2) Aras doesn’t offer a real open source product in any case.

226%: Since Aras doesn’t publish any figures, we must make major assumptions to conclude they’re a competitive threat to providers of traditionally-licensed PLM. Are these assumptions realistic? By extrapolating sales volume from Aras’ published download lists and executives’ comments, we can guess that last year’s first half figures are closer to $1M than $10M, and so the press release might mean Aras increased sales by another $1M or $2M. Compare this with, say, PTC, which in the same period increased new Windchill license sales by $36.5M. Now add PTC services & support revenue, which should grow in lockstep. PTC, for one, should not be terribly concerned.

Real OSS: I think it’s fair to characterize the Aras license model as a hybrid closed-/open-source, and the sales model as shareware (download to use it “as is”, opt-in for support & security patches). The product core, the Innovator framework, is both closed source and required, while the community complement is both open source and optional. Marc Lund, VP of Aras, has said this chart did “a particularly good job of characterizing” the Aras closed/open model: http://news.cnet.com/8301-13505_3-10244853-16.html

The reason Aras can offer Innovator at all is not because their OSS community developed it cooperatively (a la Linux), but because their original investors spent millions funding a traditionally-licensed PLM software to compete with Agile. As Marc Lund once noted, the original Innovator PLM sold relatively few copies per quarter and Aras needed a new strategy. “Enterprise open source” was born. The original Aras PLM system was easily segmented into the proprietary Innovator framework and community-developed JavaScript and XML text files.

Innovator is covered by the unusually broad Aras Innovator Agreement 3.0. This includes not just the usual closed-source stuff (“…may not reverse engineer, decompile, or disassemble the Licensed Software…”) but also inserts conditions on the optional service subscription (minimum two year commitment, renewal at list price, 90 day notice to cancel, etc.). Interestingly, the license further stipulates that any plug-ins developed by an Aras partner are incorporated into the Innovator closed license. You won’t find these terms in any open source license agreement.

According to Aras’ marketing, the Innovator license, when coupled with community plug-ins published under an OSS license, makes Innovator “enterprise open source”. However, Aras is trying to redefine “enterprise open source” as a closed-source core with open source optional components. (Google the term, and you won’t see much acceptance of this definition outside of Aras.)

To starkly out it: if Aras goes out of business, the Innovator code can’t be maintained by the community, ported to another database, use a different client or run on anything other than Windows. The licensing is closer to Microsoft Excel with community-provided spreadsheets, and pretty far from Linux, JBoss or Eclipse.

Oleg, if the Aras hybrid license/shareware sales model is a competitive threat to traditional PLM suppliers, that could be an interesting discussion. There are relatively few successful shareware companies, and perhaps Aras has cracked the code.

But I can’t agree with using them as the benchmark for how OSS may affect the PLM market. Building a “community” around a proprietary framework misses all points in favor of OSS.

Aras didn’t pay me for this post. 🙂

Kind regards.

Ed,

Thanks for the vote of confidence in our ability to ‘crack the code’ 🙂 You make some interesting points, albeit, often misleading. You’re point about the basis of our growth trajectory is valid. It comes off admittedly smaller numbers than PTC. However, that doesn’t detract from the fact that our business is growing rapidly and that adoption is even more pervasive because there are many companies that use the software and don’t pay, especially in emerging economies. As a private company we don’t publish our financials, but we do run our business in a professional manner including conformance to GAAP accounting rules. Those figures represent the continuing acceleration of companies around the world paying for our services.

The comments about our combining multiple formats and therefore not “real open source” are spoken like a true FOSS zealot… it does make me wonder why if you believe so strongly you don’t open source you’re company’s proprietary PDM product at PDxpert? The comments somehow ring hollow when they’re ‘do as I say, not as I do’.

As you correctly point out our solutions are open source (OSI licenses, source available, etc). As for our framework, your description is inaccurate. The source is available to subscribers, a defined community of users. The term commonly used for this is Community Source http://en.wikipedia.org/wiki/Community_source. It’s because we combine of these two formats that we use the term ‘enterprise open source’. We may be some of the first to use this language, however, the structure is so well suited for corporate users that I can’t imagine we’ll be the last. And it seems like the analysts agree http://www.the451group.com/report_view/report_view.php?entity_id=63442 (logon needed for this one).

MarcL

P.S. That’s Marc “Lind” 😉

I’m sure that every software supplier frequently assesses alternative license & sales models. In a few cases, OSS may offer significant potential. However, PLM does not hold much promise for an OSS success story, as has been proven several times already.

PLM requires deep technical and domain expertise, but today addresses a rather limited market. In comparison, Linux enjoys a very broad audience while Eclipse was initially funded by IBM and is maintained by software engineers for their own community. The closest parallels I can identify are Sugar CRM (not a conceptually difficult product), and OpenBravo ERP, which has a much wider constituency than PLM, while the closed-source ERP choices may be more hated than loved. Even so, all OSS revenues remain tiny compared to traditionally-licensed software.

Market acceptance of PLM OSS seems more related to the “free, as in beer” idea, and not because OSS will provide a superior PLM experience. OSS projects must rely on the community for development resources, and will necessarily be limited to that community’s skills, time and motivation. In a narrow market like PLM, only well-funded, deeply-knowledgeable developers can explore innovative GUIs, assess different databases and operating systems, and experiment with radical data management approaches. OSS PLM will unavoidably lag better-funded alternatives.

Commercial OSS (“COSS”) like SugarCRM is often provided in two fully-functional versions: feature-light and free; or feature-rich and paid. The supplier’s incentive is to provide a good quality experience in both versions, with the hope you’ll want to pay for the richer features.

The Aras closed-core/open-complement model poses unique problems for both supplier and customer. In the CC/OC scheme, everyone downloads the same core and plug-ins, but only the service subscribers get bug fixes and security patches. The quality and stability of the code – not the features – is the differentiator.

Obviously, neither the COSS nor CC/OC supplier can survive if non-subscribers enjoy the same software as paying subscribers. Clearly, the COSS guy shouldn’t add great new features to the free version. More seriously, the CC/OC vendor must avoid publishing core enhancements (including bug fixes and security patches), since each new release re-aligns the non-subscribers with the subscribers.

I’m sure the Innovator framework can (slowly) evolve despite this inherent conflict. However, the longer the conflict remains, the more difficult it will be to offer a competitive feature set. I agree with you that Aras will ultimately become real OSS, but only because there’s no other alternative to resolving this conflict.

Kind regards.

Ed, I agree with your identification of challenges in providing of free vs. paid/subscribed versions for commercials OSS. This is different from Linux is similar systems. However, the knowledge barrier and relatively small market is the top inhibitor. In my view, OSS stories are all different and there is no main-road-scenario how to build an open source project. In many cases OSS was funded by commercial companies (i.e. IBM, HP etc.). Aras story contains a sequence of events, and maybe these events can generate PLM OSS success. I don’t think we have answers, for the moment. The real challenge for Aras is OSS foundation and community building. In parallel, I can see Aras pseudo-OSS model as something that can generate future interest in PLM OSS projects. Just my opinion, of course… Best, Oleg

Oleg – Have you seen the source code for Aras? I’ve looked hard for it all over the internet and had no luck with finding the source code for Aras. You have any idea?