If I were to ask you what the world’s most largest open source firm was, how would you answer? Red Hat, perhaps, because they’re the first pure play to get within hailing distance of a billion in revenue? MySQL, because they were valued at a billion? IBM? Sun?
The obvious definitional challenges of the question aside, I would personally pick none of the above. Far and away the biggest open source business that I’ve seen is Google, which went from $0 to a $181B with a B market cap in 10 years. True, they’re not really in the business of creating open source software – their project contributions aside. And true, they don’t sell open source software, at least in the traditional sense.
But there can be no debate that Google is a firm built on and around open source software. That their business model has little or nothing to do with the software retail business means little, as far as I’m concerned. They don’t sell their number one product – search – either. They’ve instead discerned the single most efficient way to make money using open source software that I’ve seen. Why should I care whether it’s “using” or “selling?”
Which brings us to the current debate, begun by Savio, continued by the 451’s Matthew Aslett, and back to Savio. The question being asked is one that’s made the rounds before, and one that I’ve asked myself: are there artificial limits to the revenue potential of open source firms? As Aslett puts it, a glass ceiling?
Once upon a time, I was reasonably aligned with Savio, whose position is not particularly ambiguous:
Yes, OSS lowers marketing, distribution, and sales costs. And yes, OSS is a great way to drive revenue from $0 to $X. The value of $X differs based on the software segment in question (operating systems vs. business integration vs. databases vs. content management). I’d put the value of $X at $100M for most software segments. Once an OSS vendor approaches $X, their business dynamics, and more importantly, customer dynamics, change dramatically. These changes were not wholly understood by OSS proponents making the “repent or perish” claims, simply because virtually no OSS vendor had run into their $X figure at the time.
As the vendor reaches $X, they have saturated Category “C” users (those with cash and willing to spend it to save time). Now, the OSS vendor must try to win with Category “B” users (those with cash, but who have been trained by the OSS community to expect value for free). This is no small task. It is, however, a task that requires significant marketing and sales expenditures. The only way that you can convince these users to pay is through the same route that proprietary vendors have been using for decades; sell proprietary products. Sounds vulgar, I know. I’m sure many OSS vendors will try novel tactics instead. It won’t work. Selling proprietary products, while sure to draw fire from “the community”, is truly the best way forward.
Here’s how I put it back in 2006:
With all that said, however, do I see MySQL becoming a billion dollar revenue machine in the kind of timeframe that I would term as “soon?” No. MySQL, and most other open source software players, simply do not command the margins that their closed source predecessors did which more or less mandates a volume play. Volume, as I’ve said, is the surest path to success but nobody said anything about it being the quickest.
Over the past two years, however, I’ve reconsidered just what it means to be an “open source business,” and I’d argue now that the definition that Savio is using is artificially limiting. Much can be learned, I believe, from Google.
Obviously, there’s but one Google, just as there was but one Microsoft before it, one IBM before that, and so on. But the question isn’t whether open source firms can become Google – they can’t – but whether or not the lessons that Google represents are applicable. I believe they are. And said as much:
The commercial open source world can be seen in that context as one of the larger and more interesting public economic experiments imaginable. What will people pay for? Where can commercial entities add the most value? All these questions and more are asked and answered, after a fashion, on a daily basis.
What is increasingly clear to me, however, is that the traditional support and service models will be augmented – and perhaps replaced, in some cases – by network services and offerings.
Much of the revenue directed open source firms way thus far has been related to their willingness to pick up the phone when things go awry. In that respect, it’s remarkably like insurance. And though insurance is a fine, profitable business to be in, it’s not likely to be quite as rewarding or profitable in the software business as it is elsewhere. Nor are the ISVs needs aligned particularly well with the customer. If the ISV does a perfect job with the software, after all, why would the customer pay for support?
Network services, for the most part, do not suffer from these failings (though they may, of course, suffer from others). If customers share some data and telemetry back with providers, both parties may benefit. And that service will prove to be more compelling, I believe, for customers skeptical of the value to traditional support and service.
Much of the current discussion centers around the difficulties open source firms have functioning as traditional retail software vendors; little consideration is given to the possibility that some percentage of open source vendors could escape the assumed revenue ceiling by becoming something other than a traditional retail software vendor. One that, like Google, has little if anything to do with traditional licensing, maintenance, or support/service revenues – either open source or proprietary. A hybrid open source/virtualization/cloud offering? Why not? Or maybe it’s a business like SpiceWorks that recoups its costs via an ad network. Perhaps it’s one that derives value from the “wisdom of the crowds” like Splunk with SplunkBase. It could even be a creative usage of pattern recognition on incoming telemetry, like Google’s “Did You Mean?” Closer to home, MySQL Enterprise’s Monitoring & Advisory Services are a hint, and all three of Canonical’s Landscape, Red Hat’s Exchange, and SuSE Studio could be potential vectors for precisely the sorts of network services I anticipate underpinning value.
If it seems as if I’m downplaying the significance of the software here, that’s because I am. Because while software is always the foundation, it’s entirely possible that the value of the source code from a retail standpoint could be far less than the value of the data that it’s used to collect, as Nick Carr argues. If that bold assertion proves to be true, or even partly so, we may yet come to regard these periodic arguments over open source business models as quaint and archaic holdovers from legacy monetization strategies.
In any event, I remain convinced that it’s by this mechanism that open source vendors will avoid the revenue plateaus that many conclude are their inevitable fate. Some may argue that I’m cheating by blurring the definition around what constitutes an open source vendor such that Google qualifies, but I’m fine with that.
I’m pragmatic at heart, and am inclined to take the shortest path to a 180+ billion valuation. But maybe that’s just me.
Disclaimer: Of the mentioned firms, Canonical, IBM, Microsoft, Red Hat, Spiceworks, Splunk, and Sun (MySQL) are RedMonk customers, while Google and Novell (SuSE) are not.