How to Create a Profitable Desktop Business for Linux

I've been involved in Linux since the early 1990's, worked at one of the mainline Linux distros some years ago and have been an Open Source author of two projects. Over the years, the Linux environment has made great functional progress, yet mainline Linux vendors have struggled continuously to create a profitable desktop business around it. Many have tried, some have just given up on the desktop.

Given much of the software in the Linux environment is free, it seems a natural corollary that a Linux desktop would not be a profitable proposition. But I contend that not only can Linux be a profitable business, but it will be. Making money selling a shrink-wrapped Linux OS, or ISV apps, I'd agree is a very tough gig. But a number of factors are aligning to make an entirely different kind of Linux desktop business a very viable proposition. The equation:

profitability = build-ecosystem + app-store + reoccurring-revenue

By "build ecosystem" I mean proliferate a given Linux distro as widely as possible. Give it away if needed. Do whatever it takes to get it on as many platforms as possible. It's seemingly antithetical, but one can not be concerned with making profit here. Spend money to make this happen.

Build out a low-friction (as in brain dead easy to use) app-store, which users can click-and-buy to get new exciting apps. Put a lot of energy into this app-store -- it's the foundation of why your distro will be profitable. Build gravity and critical mass. Make developers eager to get their apps on your app-store. A percentage of apps being free is a good thing, on the app-store -- build critical mass like your life depends on it.

Reoccurring revenue. Linux vendors need to get out of the mentality of selling Linux. Rather, Linux can be a platform and a vehicle for generating after-market and reoccurring revenue. Some concrete examples follow.

Does this sound a lot like Google's (GOOG) Android model? Yep. And as far as reports go thus far, they don't even seem to be concerned with making money on the Android Market itself. Android is such a great vehicle for mobile search, location based services, content delivery etc -- apparently that's plenty. And by the way, Android is founded on a Linux kernel, and there's a lot of buzz about it pushing up into netbooks and higher food-chain devices. I would think Android will end up being a highly profitable proposition for Google, just not in the conventional Linux distro business type of way. That's the mentality I believe will make a Linux desktop profitable.

As promised, here are some apps that would support a "back end" loaded Linux desktop business model:

• Backup to the cloud (user backup is a thing of the past)
• Pay-to-enable hardware features widget (take cut of enablement)
• Multi-media search redirects and sales (Boxee, Roku, ...)
• Personal navigation / mapping software & database (Garmin, ...)
• Smartphone companion software (e.g. Celio)
• Pre-installed Windows/Office VM
• Voice recognition software (plugs into a host of other potential services)
• Location based services redirects
• Android VM (revenue share with Google?)
• 3G service enablement, pay-as-you-go, or multi-vendor trials (kick-back)
• Other apps on store (take % for commercial apps)

I can think of a few more money makers, but hey, not all my apps are free.

Comments

[Justin Cred...:]

1 problem I see with this. You need cheaper hardware to be competitive. Users won't choose an expensive PC with Linux over Windows, the cheaper hardware will fatten up margins and performance will still be great, cause we all know you don't need 2 gb ram, 3ghz cpu, and a 256 pci express video card to run Unbuntu, BSD, Suse, etc. But who will make the cheaper hardware, Intel or AMD? Good luck with that. Bottom line, lower hardware costs if you wish to see open source gain market share. I.E. 1 Ghz cpu, 256 SD Ram, 20 GB hard drive, 17' crt, Unbuntu Linux, fully loaded with apps like open office, mplayer, etc. $149-$199.

Jan 14 10:25 AM

[bigmoney:]

I don't know. I've been a Linux fan since early 90's as well. Linux has a steeper learning curve than Windows, that many people aren't willing to bridge, esp. when Windows and Windows apps are either heavily subsidized, or else pirated. Who buys copies of Windows and Office for home use? Just get an enterprise license key from someone you know, and press on.

Some setups may be suitable for the app-store model. Perhaps appliances. But the problem domain for linux and open source is huge. Who has any idea what plugs-ins are available for Firefox? Or what all is out there on sourceforge / freshmeat? In the linux world, whatever you want, chances are its available and free. It may not be easy to use, or very reliable, but its probably there, and waiting for a programmer to improve upon it.

There are some applications that Linux just doesn't do well, but those are often limited by business reasons. Hardware vendors won't release specs, video copyright, proprietary fonts, print drivers, etc. Or perhaps no open source coder wants to produce a high quality Quicken Turbotax. Who wants to mess with the motherload of tedious and boring, but rapidly changing tax code if they aren't getting paid?

Also, alot of the best open source programs are available for Windows. So many folks, myself included, find themselves using Firefox, and the suite of PortableApps.com on a Windows platform just out of laziness.

At present, I run two machines, one windows and one linux. Each does things better than the other.

Jan 14 12:45 PM

[Chris B:]

"By "build ecosystem" I mean proliferate a given Linux distro as widely as possible. Give it away if needed. Do whatever it takes to get it on as many platforms as possible."

This is where the product itself, rather than the distribution model, is important. Windows users just aren't willing to endure the added complexity and frustration that the more-secure Linux imposes on them. Mandriva has done a lot to create a secure but Windows-like environment, but users still face the following problems:

-Finding drivers for devices.
-Finding codecs for common but patented media formats like mp3 and m4a!
-Programs like flash and java not included in distributions.
-Confusing file architecture and naming. What goes in bin or etc? Program Files?
-Installing software without resorting to 1970's command-line. Not everyone programs for a living.

The Linux company that fixes these few minor problems will grow explosively. Read the forums and message boards - people frustrated with Windows try distribution after distribution, looking for user-friendliness and not finding it.

Jan 14 03:59 PM

[jack dee:]

App store, wow that is such a new idea, great. All you need now is a product /Brand and few thousand staff. THis is all so fresh and new.

Jan 16 12:22 PM