Adoption Chasm

If like me you were paying close attention to Mark Shuttleworth’s key note introduction to UDS you may have noticed he used an adoption graph like this one:

The idea is that right now we’re stuck between 15% market share of early adopters and the magic 20% market share of the early majority… after which the adoption momentum will be a self sustaining. What I took away from this graph was that there are a spectrum of people in the world who progressively need great numbers of friends to already be familiar with the product/idea and the more happy people the more of the majority can be won over.

But hang on a second we need to know what we’re talking about first. If we’re talking about computer desktops then Ubuntu (and all other distros together) could barely be said to be 3% of the general population market share, far short of the 15% of the adoption chasm that Mark was targeting.

Of course we might be talking about a subset market, maybe netbooks where I think Ubuntu does better? Perhaps we’re at the chasm of early adopters, perhaps we have 15% of the market of Innovators and Early Adopters (2.25%) maybe going after the general majority shouldn’t be attempted until we’ve conquered even a significant proportion of the early adopters in this model.

That would mean that we need to stop serving the innovators (the programmers in this case) and start pushing early adopters, people who will put up with a bit of grief but will if they fail forever label the brand as a reject (and will be sure to tell the early majority it’s to be avoided).

Perhaps this is what I find when I go out into the street, the early adopters know Linux is failure, it didn’t work when they tried it 10 years ago and it’s certainly not going to work now. On the other hand Ubuntu is new, a fresh brand… it doesn’t help perhaps that there is a lot of marketing promoting the link between Linux and Ubuntu, dragging Ubuntu down.

Then again perhaps Ubuntu really is that crap that it should be rejected. There’s a whole pile of support questions in my email archive that are from frustrated early adopters who can’t get their printer working, their scanner, ipod, wacom tablet. Or perhaps it was that website, that game or something else. A computer is a very many to very many device and we’ve decided we’re audacious enough to attempt to service the greater proportion of those combinations without any very many to very many organisational procedures to test any of the combinations above casual testing.

I’ve heard that some have said Ubuntu won’t take off until we’ve stripped away some of the excessive expectations and slimmed down what is possible to do until we have a manageable subset. Perhaps netbooks and similar small devices are an aim to do just that. I believe that Canonical knows this well enough and is preparing a solid system for a very small set of functionalities which can be supported by an individual company.

Perhaps it’s up to the community to bring in the rest of the possibilities that we’ve vested ourselves to.

Your thoughts?

14 thoughts on “Adoption Chasm

  1. I’d hate to say it, but I think we will hit the tipping point once several things happen:

    1. Adobe publishes their Adobe Suite on Linux. They already have started flash (which still isn’t that great and is buggy on amd64) and reader, so all hope may not be lost.
    2. Major game publishers start making linux clients. This too seems like there’s good progress being made especially with Valve porting Steam over.
    3. When Ubuntu OS focuses more on people. For example, desktop sharing (done, but needs better resolution), multiple video conferencing (jingle?), meta contacts (work is being done with libfolks). WebM looks like it could help this in many regards.
    4. Having some better applications. Have a much better and less buggy photo management tool than F-Spot, a better file system (btrfs) with some user friendly tools (think Drobo), some better theme tools and better themes.
    5. Lastly, they’re making good progress, but the UI needs some work and I think they need to do something (read get rid of it or put it somewhere else) about the menu bar (File, Edit, View, etc).

  2. The various FOSS communities have done incredible work on our software. We have momentum, immense talent and great tools with which to do our thing.

    Obviously when we have more compelling applications which are found only on Linux, then more early adopters will be more interested in putting up with the rough edges. But these applications are totally dependent on the underlying technologies. So I really do think everyone’s efforts are valuable.

    People who last tried Linux years ago and gave up may not be worth worrying about too much. There will be plenty of youngsters to take their place.

    As for producing a stripped down effort vs the more difficult goal of a general desktop, I think the latter is valuable because producing applications for it means pulling in a greater range of developers.

    One more thing: we apparently need more applications with spell-checkers — check your post’s title 😉

  3. I’ve said it before — I’ll say it again: Ubuntu needs to concentrate on becoming a platform, not “some Linux.” It needs to slim down, ejecting a lot of the stuff from universe and multiverse so that it can deliver a “truly works” product, create great, integrated developer tools (Ground Control, Acire, and Quickly are helping), and make it easy for users to find and add PPAs (which cover the software that Ubuntu ejected and which won’t be associated with Ubuntu if the software doesn’t live up to expectations). Bonus for adding a “pay” option for PPAs to attract devs.

    Linux is truly just a kernel. I know that sounds like GNU stuff, but I mean that Ubuntu is an OS — a platform — and Linux should have little to do with that in marketing or people’s impressions of Ubuntu.

  4. I run a small computer repair business and had been running Ubuntu 8.04 Desktop on a Dell Laptop, 8.04 Server on a ‘white box’ server and Ubuntu 9.10 remix on a Dell Mini. The 8.04 environment worked very well. All have been upgraded to 10.04, with the Mini getting the remix.

    The server and mini upgrades were terrific and problem free. The laptop was a disaster. (The Flash-nonfree problem wrecked update manager.) I finally landed up throwing in a new hard drive, partitioning it to ext4 and loading 10.04 from burned CD-R. It’s taken me over a week to get logins for ssh, vnc, rsync, all my files back in their usual directories, e-mail, etc., etc. back to normal. Wine still isn’t okay with the password manager program I normally use. Yuk! (I don’t like the Linux password manager programs I’ve seen.)

    I’m a techie in the Windows and Linux world and don’t really mind too much about hardware/software issues. I just get on with dealing with them or using a work-around. However, the new environments though slick are very hard to get used to. (I’m 70 years old.) From a business efficiency point of view, the upgrade was far too problematic and it interrupted my business work while I chased down almost endless preference issues. There simply are too many feature changes from 8.04 to 10.04. I installed Gimp which I consider necessary for my webpage & website work. I’m sure I’m missing some other programs I use from time-to-time.

    At the moment, I wouldn’t recommend anything but a clean install of a new Ubuntu distribution from a CD-R that’s had its md5sum checked. Most businesses wouldn’t put up with what I went through.

  5. Even today, Ubuntu is not working as it should on my laptops when it comes to plugging in projectors.
    I’m a lot on conferences and meetings and there is the trend that Windows just works with projectors, while with Macs there is some work to do after you plugged it in but with Linux (mostly Ubuntu) it often does not work.
    That’s very sad, but this seems to be a major reason, why many of my colleges left Linux and bought a Mac (we all want some kind of Unix). I guess they have been early adopters…

  6. @Eric Pritchett

    Why do people like you put so much emphasis on Adobe apps? Flash aside, Adobe has nothing to do with Linux adoption. For everyone who cares about the Creative Suite there are another 100 general computer users who will never go near it in their life.

    Yeah it would have helped if GIMP hadn’t been in denial about its problems for years and had switched to a single window interface years ago, but really, a lack of pro level apps is not really a driving force in mainstream adoption.

  7. Ubuntu (& linux in general) needs to get some very basic things right, which until now they have failed to do.

    Video support – Ubuntu still has huge issues running dual screens, which includes projectors. Every Ubuntu distro over the last two years has had significant issues with Intel, Nvidia or ATI drivers, and 10.04 has some serious bugs with Nvidia and multi screen that have recently been put off to 10.10 (at least). Linux’s poor video support also underlies a lot of the gaming future of the platform.

    Sound – is only just starting to come into some sort of sane order in Ubuntu 10.04 – this is so basic, and yet it’s been a long time coming. There needs to be one unified sound architecture, or if there have to be several they need to at least play nicely together.

    USB – several system crashing bugs *still exist* in the linux kernel drivers for USB 2. There are bug reports going back 5 years on this issue alone, and they are falling on deaf ears. Don’t believe me, do a google on the words ‘USB high speed device reset ubuntu’ (without the ‘ ‘s obviously). This bug alone turned me off the platform (sadly) because my notebook’s uptime was measured in half hour blocks at best. The only solution is to force the whole system back to USB 1, which is in itself not a task for the faint hearted, if you could stand the performance hit.

    Ubuntu 10.04 is finally getting the software installation thing better, but it still has a way to go. It should be dead easy to install a ppa and import the key, but it’s not as easy now as it should be. One day.

    Power management is also finally getting better, thank goodness.

    Printing is still in pretty bad shape if you have a printer that isn’t Postscript based. PCL based printers are *really slow* if you print anything with graphics. Hardware support for multiple copies of a document is non-existent for many printers, meaning that each page has to be sent separately, rather than caching the page in the printer’s memory once and leaving it to the hardware to handle it. Duplexing is hit and miss. I wrote to the CUPS people about how to fix both multiple copies and duplexing for my printer’s PPD and they stated (almost verbatim) that they didn’t want to know.

    Funny, but all of these things are so basic, and yet none of them qualify for the “paper cuts” push. They are so low down the priority list it’s not funny.

    I have worked in tech support and consider myself an early adopter, and I tried Ubuntu solidly for two years but sadly several of these basic things have done Ubuntu in for me for the time being. Most of them are still live issues in 10.04, with some progress on installing apps and sound.

  8. Heya Doc,

    Do you have a larger version of that image? If so, would you mind emailing it to me?

    Thanks, loved the post!

  9. The one issue with ‘crossing’ the chasm is that very often you find yourself losing the population that got you to it. The needs of the early adopters and such are rarely the same as the Early Majority. I would recommend people to read the book that started this (and probably later ones like Tipping point and such)

  10. The truth (afaik) looks like this:

    1. So far as Ubuntu is dependent on Linux, it must suffer until Linux itself is reorganized. The problem is that Linux is an experiment in progress. What little direction it does have keeps changing, and at a fairly quick pace. If there were some kind of consistent interface layer for the kernel so that each rebuild of the kernel doesn’t require a rebuild or even rewrite of drivers and applications, then maybe 3rd party companies would be more willing to target Linux, and thus Ubuntu by extension.

    2. Non-existent, partial, or poorly-supported drivers in Linux are just a pain. Take for instance what I’ve gone through just trying to get my Genius Mousepen to work after upgrading to 10.04. While the pen works just fine, the mouse is out of commission. I had it all working before. Why should an “upgrade” break what already works?

    3. Inconsistent interfaces are the bane of any system from an end user’s standpoint. Take sound support for instance. First there was OSS (now defunct), but now there’s ALSA, complete with OSS emulation. But you don’t want to code directly for either of those since you’ll have to contend with one of a myriad of sound servers just to open sound support (esd, artsd, port audio, jack, etc). I don’t mind the back ends, but can’t we settle on 1 sound server, add all needed features to that, and just run with it?

    4. Documentation!!! And I don’t mean MAN pages. Exactly why is there no decent GUI tool to display INFO and MAN pages anyway? It’s been how many years now since so-called “desktop” distributions started?

    If Linux as a whole, or any 1 distribution can manage to make at least this much happen, then it’s possible to get those early adopters to opt in. But we can’t just stop there. There’s so many other issues to be resolved.

  11. @Jimbo

    I don’t use the Adobe Suite, but in terms of early adopters I have found that the one thing holding a lot of people back are games and adobe software (not just photoshop) and no GIMP is not exactly a replacement (just look at the number of brush strokes). Could Ubuntu reach the tipping point without the Adobe suite? Maybe, but the open source alternatives would need a lot more work (a lot more). I wish this wasn’t the case, but I usually find designers closer to early adopters than normal users.

  12. Co-sign all the comments above but I would add this…

    Documentation and education. Until “I’m taking my first computer class” does NOT by definition mean “I’m learning Microsoft Windows and Office”, we’re going to be stuck at the single digit adoption rate.

    I’ve worked in adult education for several years. Virtually every college, library, public/private school and government computer skills program defaults to the Microsoft platform.

    MS GIVES AWAY their windows and office curriculum – in 21 different languages.

    Goodwill’s free training offerings – same story.
    Goodwill International

    Hell, even APPLE offers classes at the genius store.

    I love FOSS. But FOSS is not easy to adopt on any level. It takes work and effort to achieve a level of proficiency. And I don’t mean having to search through blogs, irc channels or forums. The new Ubuntu Manual is (FINALLY!) a welcome step in the right direction. But WAY more needs to be done.

  13. I think it’s worth it to realize how difficult Ubuntu Linux was to use just TWO YEARS AGO. It was by no means TOO difficult, but that’s when touching xorg.conf was an occasional occurrence, you had to download and load GPG keys, and going to for a half hour or so before using your system was common, unless you liked the Human theme. (lol)

    Linux has always had great functionality, but only recently has the design of most mainstream Linux programs become glamorous AND fucntional, as well as the ease of use going from a little harder than Windows in some cases to easier than Windows in all ways.

    No more xorg editing on pretty much all hardware I’ve touched, just add sources from a link in the beautifully executed software center… it’s all so much better, even only after two years.

    So, you must recognize that, just because it’s good enough right now doesn’t mean it’ll be adopted over night. It might be perfect for everyone, and it may even get tons of advertising on TV, and it would still take about a year before you see a substantial change in user base. So, I think we should just keep supporting older hardware as much as possible, refine what we have, and get more ISV support (hopefully we can get Adobe on board).

    Let me put it this way- most computer noobs think Macs are awesome, but I can’t afford one, so I’ll put up with the viruses, slowness, compatibility issues in Windows. This reinforces the idea that you can only get something better by paying more.

    If we had Adobe CS and iTunes already, then we would have plenty more users. When people here VirtualBox and Wine, they start to think, “this is more complicated than what I have now, it couldn’t possibly be worth it.” However, Adobe CS is by far the most requested application in the community, today. You see a lot of people dual-booting just for it.

    Anyway, getting a bit TL:DR, here. We just need to appreciate what is good about Ubuntu and iron out the kinks, and get more support- don’t let a low user base now make you think there aren’t millions of people who would use it if they only knew about it/how to install it.

Comments are closed.