Correct me if I’m wrong but never worked for Red Hat, but you seem willing to editorialize their previous business decisions just fine in spite of that fact (comment 47). As I’ve never worked for Canonical nor Red Hat, I’m equality qualified to talk about both. If you don’t feel comfortable editorializing Canonical’s performance to date..don’t…that’s not anything new..it’s not like the current Canonical employees like talking about it in detail either. But don’t try to speak to Red Hat’s business decisions either unless you are prepared to do that comparison.

-jef

Just a little historic comparison. If you start counting Red Hat as beginning in 1995 when the corporate entity “Red Hat Software” was created. It took Red Hat about 6 years to get to its first profitable quarter in 2001. It’s really unfortunate that Google trends didn’t exist for that 6 year span. There’s no way to really support the thesis that Red Hat gave up the brand awareness when RHL was discontinued that you think Google trends is suggesting Ubuntu earned in the last 4 years. And certainly the 2004 Google trends data for redhat and linux don’t make a strong case that “redhat” == “linux” in whatever sense Google trends calculates brand value. Your thesis may be valid, but the Google Trends data doesn’t support it…there’s no data set that does. We just don’t have the data record that extends back far enough to capture the critical period prior to 2004 to understand the rise and fall the red hat as a brand in the Google Trend sense.

Anyways, the history is a little messy because of the RHT IPO and the number of acquisitions Red Hat did leading up to 2001. The Canonical history for the first 5 years is vastly different in that there have been no large scale expenses in the form of corporate acquisitions and there’s been no IPO. If anything Canonical should have an easier time of it than Red Hat did building a business assuming your thesis is right and Red Hat gave up valuable brand perception after doing the hard work to build it. Canonical should have a really easy time capitalizing on that, especially with no major acquisitions to drive expenses up on the balance sheets. 6 years. That’s what it took Red Hat. Canonical should be able to hit profitability in the same timeframe surely. I guess we will know in a year.

But lets look a little deeper than that. When Red Hat introduced its rhn service in 2001, it took less than three years for rhn to have 1 million systems subscribed to it according to the historical record: http://www.redhat.com/about/companyprofile/history/
This July will mark the 2 year mark for Canonical’s landscape service. Care to make a friendly wager as to whether Canonical will reach the 1 million registered system mark for landscape in the same amount of time it took in Red Hat’s rhn? If it happens I’ll buy you a beer and teach you how to curl (just get your ass to our end of the season bonspiel at my curling club..I can’t afford your plane ticket.) Wait hold on, I’ll teach you how to curl _then_ I’ll buy you a beer.

I’d love to compare Canonical’s training services program to Red Hat’s…but Canonical is just really getting started on that. In about 3 years Red Hat had more than 6,000 certified RHCEs…if you believe the press clipping:
http://www.redhat.com/about/presscenter/2002/press_rhct.html
“Started in January 1999, RHCE is a performance-based training and certification program and de facto standard Linux certification, with over 11,800 RHCE exams taken and 6,600 RHCEs certified. RHCE is the top-rated IT certification overall for quality, chosen number one in multiple categories including overall quality of program, education, and exams. (Certification Magazine/Fairfield Research, Jan. 2002)”

Three years to become top rated cert program in multiple categories. Tick-tock/Tick-tock. I guess Canonical formally begane their Ubuntu professional training last year right?

Then there’s Canonical’s ISV services…the partner repository and associated packaging services. The service, introduced for dapper and which was suppose to “hit the ground running” with the release of Hardy. It would be very ungentlemanly of me to go into detail on how badly that is fairing as as service offering by counting the number of applications in the partner repository for any release of Ubuntu LTS or not. I might need to use two hands..maybe.

Oh wait what about channel partners? Well what about channel partners?
http://searchitchannel.techtarget.com/news/article/0,289142,sid96_gci1310488,00.html
that’s from last year. Anything significantly changed since then? Canonical’s had a year have they developed strong partner relationships that actually feed revenue back to Canonical? No idea…complete lack of transparency there. If anything Turn Key Linux is probably positioned to be a better integrator partner for VARs than Canonical is, inside the larger Ubuntu ecosystem..and that’s pretty troubling to think about as Turn Key Linux doesn’t directly support ecosystem in the same way Canonical does.

So what does that leave…well I guess OEM services. Absolutely no transparency there to talk about. Where the direct to user support services have a publicly archived pricing model..the OEM services doesn’t. The most transparency I have to go on is a statement in Ubuntu forums basically saying that the Dell pre-installs are most likely going to be a wash in terms of revenue. Okay then. System76 is finally offering Canonical support contracts to end-users for server purchases starting this month i think…but ZaReason a competing OEM isn’t. And its not clear that either one of those very Ubuntu friendly OEMs are contracting with Canonical for OEM level services. The existing OEM projects in Launchpad seem to target netbooks exclusively. Which is great and all…except latest Q1 sales reports sort of indicate the netbook market softening.
Good news about the Revo Asus nettop pc though, hopefully Canonical got paid doing some integration work for that and Asus didn’t just slap Ubuntu on it on their own.

-jef

Another one that might be interesting :
http://www.google.com/trends?q=linux%2C+ubuntu&ctab=0&geo=all&date=all&sort=0

While “linux” goes down in trends, “ubuntu” is getting close. This is somehow a combination of your graph showing all major distros with Senko’s graph (comment #15) showing all distros compared to “linux”.

Ubuntu searches have been on the rise (though the recent levelling out is intriguing) during the period of its existence, which makes sense considering what we understand (independently of Google Trends) of Ubuntu’s rise in popularity and general freshness. Consider that, as a distribution, it represents an amazing breadth of searchable properties, where MySQL might not — I bet there’s a whole bunch of people searching for both!

Finally, my points about the market are not based on Google Trends lines. You are starting to conflate a few things there, and I don’t think it helps to describe what the search trends mean (after all, I don’t think there are “arguments” to be held about the trends, just suggestions).

Check out what happens when you add “Linux” to that query, by the way. Very interesting.

Mysql.. that’s an interesting comparison. Google trends mysql and ubuntu.

Mysql is trending down! The comparison you are trying to draw doesn’t hold up. Mysql is making money..its google trend is downward…it doesn’t support the argument you are trying to make.

-jef

But this brings us back to where our opinions are more likely to diverge: Open Source business strategy.

I think Red Hat created the environment for a competitor to enter and “steal” their immense visibility and “Red Hat == Linux” status. Mark understood and very astutely described this market gap from the very beginning (before “Ubuntu” and “Canonical”).

You keep talking about Canonical struggling to make money… only five years after the very first release of Ubuntu, and only three years after the very first (business-viable) Long Term Support release. These things take time. (Plus you’re solely focusing on support revenue, but that’s a completely different conversation.)

Red Hat ceded mass-awareness and product-preference-by-osmosis — which it totally owned at the time — when it adopted the RHEL strategy, particularly because that strategy was not immediately teamed with a viable answer to the community, technologist or early adopter markets. Fedora is almost there, but I still think it gets pooh-poohed to a certain extent by Red Hat.

I think the “competition with community support” concern is a furphy… as Marten Mickos says, “time vs. money”, and people sure buy MySQL support!

So, as it will take time for a five year old company to establish itself in the mind of corporate buyers, it will also take time for Red Hat’s previous advantages to diminish.

You search for firefox plugins on google? You don’t use firefox’s internal addon tools which take you directly to addons.mozilla.org? Hmm. That’s interesting. I would have thought you and I would be representative of the same segment of the userbase… the 0.1% technically proficient tail of the population.

I never trust my own usage patterns as being representative of mainstream usage. I should do a little test, ask my wife to search for and install a firefox plugin to do a specific task and see how she does.

But you bring up a good point about “support” which goes right back to the heart of the matter. Because the Ubuntu community “support” is so strong, that undercuts the value of the “support” that Canonical is selling to fund Ubuntu as a project. Is Ubuntu as a project itself a bubble economy created by an injection of venture capital from Shuttleworth, where its own popularity is eroding fundamental financial considerations that drive the business model?

Here again the comparison for firefox is interesting. There’s no doubt that the Google search revenue is a very large financial driver for Mozilla and thus firefox development. But more interestingly for firefox, that revenue is directly tied with usage of firefox..there is a natural coupling between expanded use and the revenue associated with the Google partnership and community action doesn’t really cut into that feedback loop at all. Is that same dynamic working out with Canonical? Or is the broad availability of Ubuntu community “support” undercutting the feedback loop that would power support revenue for Canonical.

The very thing that makes Ubuntu’s brand trend in Google Trends so compelling, could be the very same thing that is disrupting Canonical’s ability to generate revenue… easy access to community “support.”

I really wish Canonical would hurry up and poop out a service offering from the Ubunet/UbuntuOne project work they are doing internally. If its anything like what I suspect it is, it’s going to get Canonical away out of the trap of competing with community for “support” and will let Canonical compete directly with other web services businesses. They have a better chance of competing there than competing directly with its own Ubuntu community as a “support” provider. A Better chance..but not a good one.

-jef

I grok your scepticism about users querying for their products, but I do tend to find myself using them all the time as query components… For instance, looking for a plugin (for Firefox), web development info (for Firefox), the solution to a hardware problem (for Ubuntu) or application availability (for Ubuntu).

But I suspect that Firefox probably has a larger proportion of acquisition-oriented searches, while Ubuntu probably has a larger proportion of support-oriented searches.

… and it’s the proportion of “acquire” vs. “support” that (I feel) is likely to have the most impact on those trends in the future.

That firefox versus Ubuntu graph is interesting. Being a trained skeptic, here’s the first set of questions it brings up in my head. Why are people searching google for the term firefox? Are most firefox queries done by firefox users? Are the being done by people looking to install firefox? What would happen if firefox was pre-installed and was the default browser would the google queries involving the firefox term go down significantly? How often do users of firefox query google for firefox versus how often non-users of firefox query google about firefox?

I use firefox. I don’t make a habit of googling for it. Do you? Do you make it a habit of googling for information about firefox, while actively using it on a daily basis?

Now replace firefox with Ubuntu in my questions and see if you come up with the same sort of intuition with regard to how to explain the querying habits. Google trends are difficult to understand without having some sort of mental model that explains querying habits. If you don’t have confidence in same mental model to explain the query habits for both the firefox and Ubuntu terms…they might not be good comparisons even though the trends are similar.

-jef