At the recent OpenStack Summit in Vancouver, Mark Shuttleworth announced that he was debating an initial public offering for Canonical Software, Ubuntu's commercial division. The news was interpreted as a sign of success in many circles, but whether making Canonical a public company would be a wise move seems doubtful at best.
As reported by Steven J. Vaughan-Nichols, Canonical has been considering the idea for several months, but has not yet made a definite decision. Yet the idea has been raised because of the success of Canonical's OpenStack consulting division, which has apparently become the first Canonical venture to become profitable, and includes partnerships with Microsoft and VMWare. "We now have a story that the market will understand," Shuttleworth is reported as saying.
Canonical as a whole has yet to be profitable after ten years in business, and has been propped up by regular injections of cash from Shuttleworth himself. In recent years, Canonical has appeared to be making concentrated efforts to become profitable, cutting operating expenditures and venturing into hardware and cloud consulting. Under these circumstances, the temptation to jump at the first good news must be almost irresistible, because Canonical desperately needs to prove itself.
However, from an investor's perspective, one success among many numerous failures remains a far from attractive record. If investors can be lined up, they could easily insist that Canonical concentrate on cloud consulting and close down some or all of its other ventures -- something that Shuttleworth is not prepared to do. Under these circumstances, the IPO could be a non-starter.
In the Middle of the Gold Rush
In favor of an IPO is the fact that OpenStack is in the same position as Linux and free software in general was in the Dot-Com boom of 1999-2001. OpenStack businesses are booming, with new ones starting up every day. IBM vice president Angel Diaz claims to have done seven billion dollars' worth of OpenStack business last year, and that is only one company among dozens. A company with name recognition that is already well-established in cloud consulting could easily take a sizeable slice of the OpenStack pie.
However, for those who do their due diligence, this atmosphere could be as much a warning sign as a means of encouragement. Experienced investors may remember that the Dot-Com boom crashed as quickly as it arose, and be wary of a repeat.
Even more importantly, the possibility of quick profits attracts immense competition. By my count, 108 companies had booths at the OpenStack Summit. Of those, some eighty were companies with no name recognition, and probably less than three years old. Possibly, cloud consulting is a large enough area of business to support all of them, but a reasonable guess is that many of these companies won't exist five years from now.
Considering that Canonical's cloud consulting is only a few years old, investors will have to ask themselves whether Canonical is among those newer businesses that will not survive. To investors with no knowledge of open source, it might well seem to fall into that category.
Yet even if Canonical is counted among the established companies, the question remains of how long it can hope to flourish. According to Wikipedia, Canonical has 500 employees. Most do not work on OpenStack, but even that number is dwarfed by Rackspace's 900, or VMware's 9200 -- and these are the smaller competitors. In contrast to giants like IBM or Cisco Systems, Canonical may be too small to have a hope of competing.
The question for analysts is whether Canonical's management software and storage solutions and its recent alliances can match those of its competitors. Shuttleworth himself appears confident that they can, but the sheer number of alternatives makes that confidence questionable. Name recognition, not software quality, could be the deciding factor in which OpenStack consulting businesses survive.
Anyway, since a comparison of all the available solutions is next to impossible because of their sheer number, potential investors may simply decide to take the safe route of supporting larger companies, particularly those like Rackspace and IBM that have more experience with OpenStack. No matter how polished Canonical's solutions, the company may simply be squeezed out by larger, more established competitors.
Too Little, Too Soon
These calculations could be upset in a instant by investors willing to gamble. Gambling is, after all, what investment in public companies is all about.
All the same, Canonical's past record and the gold rush atmosphere currently surrounding OpenStack suggest to me that the company may have difficulties attracting the investors needed for a successful IPO. A failed one, of course, would leave Canonical looking like more of a risk than ever.
For this reason, I suspect that, despite the pressure on Canonical to prove that it can become profitable, an IPO will not be happening in the next six months. The projections for Canonical's OpenStack consulting seem to be promising, but it is simply too soon -- despite the way that the company was spending money at the Summit to position itself as a leader.
If Canonical's consulting remains profitable for another year -- or better yet, if Canonical's forays into mobile devices and embedded systems can also become profitable -- then an IPO would become more plausible.
The fact that Canonical's consulting services are profitable is a major milestone, and naturally encourages Shuttleworth and his advisers to start thinking ahead. For now, however, an IPO would be a rash move that, to those who do their due diligence, is as likely to be interpreted as proof of desperation as an indicator of potential.