Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Branch: master
@rdp
560 lines (343 sloc) 27.607 kB
TODO:
apply steps to this itself
read books about it?
todo read this: https://hacks.mozilla.org/2013/05/how-to-spread-the-word-about-your-code/
90/10: last 10% takes 90% and is the 10% you don’t actually like doing [mundane], but is equally as important...
Read Mozilla hacker by Julian
.
.. Something. Very good blog post!
OK the real kicker is that you should probably not really start into
Something without realizing some caveats.
Like... You'll probably, most likely, typically, be the only leader of your project.
Maybe it would be good to therefore only stay into projects you s are
Really passionate about and could stay passionate about for years to come.
Like ten.
Because your the one who will be chasing down new features for people that you don't actually care about.
Things that don't affect you.
And doing the crappy work on it like bugs and dealing with poor input etc..
Stuff nobody else wants to ever do.
And doing marketing... Possibly all of it... And it takes forever.
At least initial marketing to popularize,
Then Google organic can kick in... Hopefully.
If it does its the saving grace of developers since in essence it does your marketing for you...
You can hope anyway :) but even with good seo maintenance will still be present and on your shoulders...
And forums and irc chat 24/7...HavingtocheckFacebook's...
And it will all be for free most likely... Our absolute beginners who know nothing yet... Training them...
You might make some money of it, or possibly convey into a
Pay project (where's the fun in that?)
But probably not. The gifted few might even make a salary.
But that is like rock star s... There can only be free...
Assume you'll possibly make nothing...
This better be a passion... Our it will end mediocre, seemingly... Our at least not polished, and not improved.
I wonder if focusing on "only one project" would help but it would need to be
Only one, for yea s... Especially when that focus contains marketing
(Which is another reason partnerships work sometimes...
One can focus on one thing while the other does the technical...)
Is either that or do both yourself which is tricky...
Especially for technical people who dont want to do much... Not natural...
One idea: try to take some more queries from lean startup
Like polling and pivoting so you can put your "decade"
With off effort into something that can become even ore popular/useful...
NB: the mentions here on google keyword stuff are less important since the EMD update: http://news.submitinme.com/newsdetails-249.aspx
so are deprecated...
plan up front of spending a lot of time (budgeting a lot of time) to marketing
NB that this doesn't include much on successful marketing...which is 80% of the product,
though getting good google SEO is a strong step in the first right direction for marketing...
Steps to make your open source project popular.
The end goal of any project should be that
people use it. And typically, you want it to
be used by the most people that would benefit from it,
if they knew it existed and used it. Basically your goal *should* be
to make your app as popular as possible.
It is worth it to try to popularize your program. It is worth the extra time and effort.
Basically, you need to spend lots of time researching marketing and sales...
Expect to spend 50% of your time doing unsultry things, like writing documentation, writing up press reports,
creating installers, etc...
also expect to spend some time (a lot?)...adding features that you don't care about (and possibly don't even really like?) because others think they'd be cool
Having an in house "cooling down time" while you use it to test it before releasing is good.
They say that a website is 80% propaganda 20% usefulness
If nobody knows your project exists, it doesn't matter how cool or useful
or better it is, nobody will use it, and nobody will participate in it.
So here are some ideas how.
1) Have a useful app.
1.5a) you can distribute it to developers before it's useful or polished.
1.5b) you should distribute it to "prototype" users during the
development process, to refine your requirements and re-assess
what people actually want.
2) Make it easy to use.
Ask for user feedback on this. And early.
3) Try not to have too many bugs which annoy users.
4) **SEO**
This is the key. Google can help you find users that you never could.
And is the great playing field leveler.
How to increase SEO:
4a) use the google keyword ad tool to see what the most popular "related" queries
are for your product. Make this the name of your project.
Ex: for a project which takes "loopback" audio and captures it as a device, the following were the statistics:
waveout mix 1600
audio sniffer 390
record what you hear 12,000
virtual audio interface 320
audio loopback 1900
record speakers 12,100
audio capture 246_000
virtual audio 135_000
audio grabber 165,000
So now choose a project name that matches the most popular:
in my case: virtual audio capture grabber device
maybe add some personal flair, like "yet another" or something to make it distinct.
rename your github project this, create a sourceforge project with the same name style, change its "project url" to point to yours,
create the "project site" as virtual.audio.capture.grabber.device.t28.net or the like. (t28.net is freely available on afraid.org)
You want as many links back to it as possible, esp. if it's a related conversation or the like.
Post to your blog a post "virtual audio capture grabber device" with a link just to your project.
Possibly several different posts with different common keyword combinations.
Also could blog like "competitor/alternative to xxx" with the name of some popular (pay) competitor et al.
Search for the various words that match (all of them, even low popularity), and try and post links back from there
to your project.
Be careful to not spam, though. Use your "real" name when making blog entries, and don't make extra blog entries, just change the url you use to your site. Be authentic/legitimate or you'll be marked as a spammer.
If you do get marked as a spammer, then change your url to somesubdomain.yourproject.com
"for the first few links, try and add yourself to them" ref: http://www.lextm.com/2008/08/promoting-open-source-projects.html
make it part of your email signature (have your name link back to it, or the like).
make a facebook group page for it, get 25 people to like it so that it can be visible or the like.
Then the secret is to "keep pushing it" over time. As people blog related things, tell them about it, et al.
You can never make it too easy for end users. Assume they'll be "way more naive" than you even think they might be.
Post to mailing lists that are related (or even once with each release or major release).
Create google group with the same name. Link appropriately.
Get your-name.com
TODO change the name earlier, rather than later.
Add yourself to alternativeto.net
Get people on facebook to "like" you or your web page
Maybe add yourself to online indexes of competitors (like dmoz or the like)
Create youtube account with name, some videos, link back to your project.
create google "project" (code) with same name, link in.
Try to have "related" sites link back to you, in their links page for example,
or have them do a blog post "about" your new site or the like. Friends, family, anybody.
If your project is a fit, advertise it on stackoverflow questions, while extolling its virtues. Same with various mailing lists "semi-old" threads or the like (don't do too many obviously).
Tweet about it. That is, be social, so that tweeting means something, then show it off.
ycombinator?
http://stackoverflow.com/questions/95710/how-do-you-promote-advertise-evangelize-your-open-source-project/7290165#7290165
has some good ideas (follow the links, too).
Basically your goal is to advertise "where your people are" (your end users). Find
them, find a way to advertise right there.
Add tracking (google analytics works) so that you can see where more traffic is doming from and how to focus perhaps).
ask your users or potential users what they would want to see
(and/or possibly "what would you pay money to have related to this project" since that's what would be more important to them).
don't have your project "home page" be github as then you can't track analytics
try to get past a tipping point artificially
Sit down with somebody and have them try it out (preferably while talking out loud). The closer to your target audience
the better. Ask for feedback (then respond to feedback and show them the fixes preferably).
In general start early, release early (but fully tested) to build your project [?]
In general start with something that "if users knew it existed, it could become popular ever" (ex "Imagine the best that could happen with your project popularity wise" to check).
You could also submit a patch to a previously existing already popular project, as a way to immediately have "your" project be popular.
If there are no google searches "near" or "close" to your idea (it's too radical or the like) then you could look for popular searches that are "near" or "related" and append them to your project name possibly, though I've never tried that.
TODO get somebody that actually has popular projects to give me the run down.
You could take over an already popular project you're interested in (or become the "3rd main developer" for it) to get some free popularity.
A few case studies: Luis Lavena took over the One click installer for ruby for windows. Its download number was already 3 million when he took over. Instant popularity!
VLC media player does something most free players don't, which is play DVD's back right (and also most files, etc.). They're by far the most popular sourceforge project.
Rails had some screencasts initially (and high-ish quality online tutorials) that really helped its popularity.
May want to include "free" in the name.
Do polls "what else would you like this to do? What about these ideas how important would you rate them?"
polish like beautiful code or documentation, examples, "book style" tutorials, books, all help.
keytool: use generic searches (two words'ish ?) and it will tell you what related popular searches are, so you can see the
best names to use.
click "order by global to see which are most popular
you can also enter "competitor's" names as search criteria and it will tell you all related searches to them.
blog posts that make it...um...match google searches that are high for it.
advertise as a competitor to that other fella :P
google the various, try to leave comments et al :)
alternativeto.net <add self as list of competitors>
present at conferences/fairs/local user groups et al (never tried this myself but it could help spread the word/look for collaborators/get early feedback)
advertise local newspapers (?) "local company creates x!"
This may be annoying, but if you want to attract "lay" members, you'll want a youtube video or something like that (with a person describing it or what not).
maybe on the main page, but don't auto-play that's just too hard.
Big "free download" links.
for code projects, and a little bit other projects as well : Until you have demos/screen casts/sample programs written in it/production apps nobody knows that they *can* show interest :)
screencasts on youtube with good "appropriate" names so that it can attract google SEO/be helpful.
google +1/facebook like :)
easy way for "lay" contributors (ex: google +)
easy feedback (feedback box).
uservoice, to allow for voting?
if you have any related projects (or perhaps any other projects at all) where you could advertise it, that could help, esp. if they're popular.
Be kind, gentle, thankful when you get user contributions. Why not?
have an contributors file, add people freely, tell them :)
"lurk" online (google alerts? stackoverflow? other related mailing lists? blog entries that get posted about
your topic?) and answer peoples' questions/point them in the right direction.
Support your product. Why not? Be nice to newbies. If newbie questions are too lame, establish a FAQ so you can
redirect them easily to something that even they can understand.
Port it to Linux if you want to have developers help you on it (and/or mac if your developer base is more on mac).
Your initial users may only be developers, anyway, so building that initial community might be good to get things started.
If there's something like rubyflow/infoq that you can submit things to, then do that.
on your website:
make it look reasonably good. At least one image, nothing ugly.
extoll virtues of your product.
extoll what makes it unique.
big download button, right on front page.
rename your product early.
don't like some other project? maybe consider fixing that project instead of starting your own.
* less maintenance burden for you.
* you avoid having to popularize your project at all.
* forking is an option, too. (and/or re-branding).
release "dev" betas earlier on, for testing, even before user betas.
pay for related newsletter, +- adwords, plead for facebook like/google +1's on your web site :)
"follow us" on facebook...
tweet/facebook it for releases/news/comments
mailing list same name as project.
domain, title same name as project. At least the subdomain.
m/b domain name like netflixinstantcompetitor.com though I'm not sure if that helps or not.
m/b subdomains with that name?
netflixinstantfreecompeitor.com
start with a project that ("imagine the ideal of what would happen with this project") could
ever become very popular.
just help some popular project instead of starting your own.
or use their libraries in your project possibly
try to advertise to "your audience"
vaguely related newsletters
real newspapers [?]
ask people first if they like your idea (people from your potential user base)
make it easy to use/human friendly
GUI/admin control panel.
Polish. Not the country. Get the fine edges worked out.
have a UI. make it friendly.
Make a "lay" version that cannot be misunderstood (expert version is ok too).
Watch people using it while they talk out loud.
look past your own needs for what it should do.
submit it for reviews by cnet, etc.
slashdot? that type of thing.
put your "name" in h1 tags, also name your movies/video with SEO friendly names
for the true work-a-holics, here's how to try and boost your facebook page more: http://searchengineland.com/7-tips-for-boosting-seo-of-your-facebook-page-91961
basically name the tabs well is all he says.
The reality is that you may need some GUI or some simplified means for them to access your program, and/or understand or use it more easily.
You need two means of contact a "feedback" form and a group. Preferably google.
Email address doesn't hurt either, though you can post the email to address of your group, too.
Don't be afraid to not do your own project, but just become the "5th man" in another popular open source project.
After awhile and many high quality patches, you'll probably become well accepted, and your contributions
make a difference, not to mention the ease of not having to take on the "whole load" of maintaining/popularizing
the project yourself, as well as possibly being able to avoid having to popularize the project at all, if it's
already popular!
Sourceforge is a good place to host your "main" site since you can have a custom subdomain url,
and add your own analytics, and have free hosting.
TODO note META DOMAINS :)
ping me if you want a subdomain under inet2.org or t28z
or hosting it under github as "github pages" works, too, subdomain wise effectiveness...though possibly a bit odd...
don't be afraid to "pivot" your project every so often, with feedback from users, so that you can create something of most value.
You want to create something of value because that would be good for society, right?
TODO instructions on "donation question post install" (link? open candy? text field?)
register with me and I'll give all the adwords profits to you! (or, give me your adwords ID...)
possibly something that connects bid for fix'ish with their checkout? ???
since your goal is to help people, in the end, become the 5th man/documentation guy/mailing list helper for some other project or group.
They'll love it, you'll make an immediate impact, *avoid duplication,* life is good.
google alerts :)
you may need a "super simplified" version of your app, for distribution, for non technical users.
The mark of a truly polished product (not that polish matters, functionality is 20%, propaganda is 80%) is that it runs "wonderfully well" on windows. Like VLC. They've polished the edges off in windows.
For software projects: documentation isn't finished until you have a book.
Once you have used your software for too long, you have no idea how normal humans will see it.
Or how they will use it. Or what will concern them. You need to run it through/past a few of them first.
screencasts for "getting started" (and/or "how cool this is") are needed.
quality over quantity
become a contributor for a lesser maintained project
maintain lower "layers" of software to have a big upper impact.
maybe advertise next to some related projects, or (if retail) list things on ebay in the hopes of attracting traffic back to your site a bit (interested parties).
"push through to perfection"
(and have something that people would use if they knew about it, I guess, but quality over quantity anyway :P)
Maybe sometimes in order for a program to be useful to people, it has to be made more system whole.
Like, not a directshow screen capture device, but something that makes screen capture easy/awesome using it.
If something is more "generally" useful (a new mouse trap) then it has the potential to become more popular.
VLC "a video player" but I suppose that doesn't prevent you from writing something small, high quality.
Find a niche, fill it...I guess... (an end niche might be better than just a library...indeed*)
Sometimes you have to have tons of value added before...you can add value, if that makes sense (angie's list).
Seems like once you have the right name, and the lurking, then you just need tons of value.
Your command line utility is not complete until it has some type of wrapper GUI on it.
It helps for web sites to add tons of "useful" data first somehow...esp. if you want to crowd source it. Like wikihow
needed/needs to pay people to do most things originally [or was it ehow], and how amazon
pays people via the first vine program to write high quality reviews. It is forcefully adding crowd sourced useful
data.
The ideal is quite possibly a "multi tiered" approach, where each tier is its own full blown distro somehow,
so that others can notice/re-use/know of the individual pieces bundled underneath. And you can get SEO
for each piece.
Developing a "final, end user" app for your individual component pieces may help them to get used.
advertise, advertise, advertise. it's what business people love, more exposure, and so should you.
I think the principle is...work on a few projects and polish them, not tons of projects that don't affect many people.
normal end users want some wrapper that insulates them from *all* of the output and complexity of the lower layers. dirt simple, your own layer on top that abstracts out all the others.
functionally complete != polished!!!
add a link to your "stuff" in your email footer. spammy, but possibly effective.
make yourself "very useful" and you will get the users (ex: firefox, but other things too)
todo: read the blog book, 37 signals, whatever blah blah about software dev.
http://www.opensiteexplorer.org "might" show you which of your backlinks are actually useful...like they don't have a rel nofollow
Actually, trying to add your site to wikipedia articles that are related "might" help with SEO, even though they don't give credit for the URL.
It "might" build traffic as well.
Better to focus on just "one thing"? (lyrics vs. sheet music?) first?
Once you're done with it, that doesn't mean it's finished.
If you can build a tool that other people can use "for their projects" then that could be useful.
Making others money is typically popular.
A tool they can use to build their dreams/help them is typically useful.
You should make it functional, and also make it pretty.
The best way to "fine tune" it to match your audience is to walk a mile in thier mocassins. Or ask them. Or if they're techy, they'll just tell you.
"In other words, the idea is that there is no way you can fix the command-line interface, for a number of reasons, not the least of which is the inertia of a precious few power-users, which according to Apple's Human Interface Guidelines is not a good thing: "If you try to design for the 20 percent of your target audience who are power users, your design may not be usable by the other 80 percent of users.". "
"Like" making it easy to use for end users. Make that your goal.
Also making it easy to install/get, easy to use.
People like things that are useful to them. New mouse trap, som tool for marketing, a DVD player. Hmm...
Really an Open source project should "pivot" to do/know what is useful for them.
github "readme" links might add some SEO. Probably not enough to matter :P
Don't rely on somebody else to make a pretty gui for your app. You make it!
Make it as good as you need/want it, then envision/go one step beyond.
("Imagine you were making it for paying customers, what features would you add? How would you change it to make it more user friendlier?")
In true, normal business sense:
take surveys to see what people would actually "like/want" to use if they could have it, too.
See what do/don't use currently, too.
Ask them for feedback for the program, i.e. what they'd like, too.
For a library, you need good "documentation coverage" of basically everything. Like examples, in the documentation/tutorial, of how to do each/everything.
ref: http://betterlogic.com/roger/2012/06/best-documentation-thoughts
Every project, well, maybe not API Libraries, needs a website (or at least some way to get to the documentation/tutorials straight-forwardly).
Non API, definitely needs its own dedicated website.
To get a language/runtime popularized, you have to do a lot of things "for" people--like a quality standard library. This same principle may apply to other things...
Put on final touches like a professional looking webpage, the "installer icon" of the installer package, etc.
Making things easier for the end user can be excellent.
Actually working and being user friendly is only 20% of being successful. Rest is propaganda or something else that I don't understand...
get awards: http://code.google.com/p/conemu-maximus5/wiki/Awards
"have popularity contests somehow" that can help
press releases?
read books on it,like acangiano's on blogging
You need a pro looking website, and pro looking installer, and pro "marketing"
something like that...
You may need to "know how" to build up a community...
real website url probably helps
advertise on alternativeto.net, SO, reddit, community forums "related to your topic" [?] etc.
you can't rely (at the get go) from crowd sourcing people to do your work/propagate things/get the ball rolling for you.
Nor can you rely on "other developers" to pitch in/join the team immediately and start adding features/fixing bugs/
making it work for you.
The only thing "end users" really end up caring about is 30% functionality, 70% the "look and feel" and "ease of use/ease of setup" LOL.
Something close to that anyway LOL.
They say building a tool will bring people running but...building something intuitive might be more important for end users?
should pivot often to meet "real" user needs...
user feedback is critical for developing what people actually want to use.
I'm not sure how closely this is related to actually becoming popular...
realize that "closed source, free" == "open source, free" to typical end users...I guess.
submit it to c.net for reviews or the like
Don't expect basically any other developers to join your core team. For some reason, it's really rare (like FFmpeg rare).
Small contributors, possibly...feedback from "lay users" more probably...
having an irc channel (with easy connection somehow) might be a nice thing... ideally inline on the website :)
expect it to take more time than you anticipate it will take (well coding always does but...for the polishing and
advertising and doing things beyond what you would normally do just to get it working for yourself to use).
this may mean that you can't do as many simultaneous/different projects at one time. Limit yourself to one or 2.
a quality "README" doesn't cut it. It should be a website explaining the concepts to users.
Market early and market hard. Since it's 80% after all...
Even before functionality has been built. It should be stable probably I suppose...
So I guess polish up early releases to a marketable state?
Don't need bells and whistles...
"real" end users only care about, mostly, the final final end product "what *could* this be, at its conclusion/fruition
with all the bells and whistles added, etc."
of course, you'll want to pivot to really know/understand what end users end up wanting...
todo: re-read http://blog.clojurewerkz.org/blog/2013/04/20/how-to-make-your-open-source-project-really-awesome/?utm_source=statuscode&utm_medium=email for fun
seems like you almost want one guy just full time doing PR/getting feedback from users/beta testing
while the other works on it...
submitting it to C|net, polishing UI [?] maybe there's enough work there...
Best way to test is to have somebody use it for an extensive period of time...(or you use it). In use you tend to run into things...
excellent testing :)
People dont' care how excellent your code or test bed is. They just care about the polished end product...like the technology side,
per se, doesn't matter...awesome code? don't care. rockin' gui? care, possibly more than I imagine...
It seems that if you do your project then are just fanatical about promoting it, watching ML's so that you can promote it,
basically getting all fanatical about it (and possibly about supporting it? adding new features to it, etc.?) Basically devoting yourself
to it, then you could make it more successful [viz: some of my ruby projects...I...just...don't care enough to promote them].
Don't be afraid to spend money (your own personal money, and possibly a kickstarter too?) on it.
Also ideally you should get customer feedback, and meet some niche that people want/demand...if you want to be popular anyway :)
Maybe in the ideal world people would make money at it?
Jump to Line
Something went wrong with that request. Please try again.