CSS Expressions. Aka Dynamic Properties.<p>Recently deprecated <a href="http://msdn.microsoft.com/en-us/library/ms537634(VS.85).aspx" rel="nofollow">http://msdn.microsoft.com/en-us/library/ms537634(VS.85).aspx</a><p>Ironically now showing up in other browsers own extensions.<p>Reasons for failure:<p>* distrust of Microsoft inventions in the browser<p>* incomplete implementation strategy based on javascript instead of custom expression parser<p>* security issues because of using javascript instead of a custom expression parser
Documentation.<p>Probable reason: some people would rather perform their own appendectomy than read a manual.<p>Seriously, I have often included features on strong user request (MUST! HAVE!) against my own better judgement, only to find that, surprise, they are not used. I don't necessarily regret it, there is a certain advantage to looking responsive regardless of the request.
I guess it'd be this entire website (visual amazon search):<p><a href="http://bigbooksearch.com/" rel="nofollow">http://bigbooksearch.com/</a><p>Reason for failure? Google doesn't index search results...
I wrote a couple browser extensions for Amazon Gold Box: <a href="http://webslices.s3.amazonaws.com/goldbox.html" rel="nofollow">http://webslices.s3.amazonaws.com/goldbox.html</a> (IE 8) and <a href="https://chrome.google.com/extensions/detail/abgkkmmoanpopajomeejifhhdgpanndd" rel="nofollow">https://chrome.google.com/extensions/detail/abgkkmmoanpopajo...</a> (Chrome)<p>I actually got the web slice accepted into the IE addons gallery (<a href="http://www.ieaddons.com" rel="nofollow">http://www.ieaddons.com</a>) after waiting a couple months for approval. Then I made a small revision to the entry, and for some reason that made it go back into the "pending approval" queue. I sent several emails to the website support contact and got no answers. Months later, Amazon came out with their own extension. I finally got a response from a Microsoft rep after escalating with a friend, and they told me about how my extension can't be posted due to a partnership agreement. Pretty lame.
<a href="http://www.thisistheonlyone.com" rel="nofollow">http://www.thisistheonlyone.com</a><p>Made a site in college to sell a T-shirt for $1,000,000. Didn't work. But I did make $40 off an MIT professor who bet me I wouldn't try it.
Bread Crumb for Emacs, <a href="http://breadcrumbemacs.sourceforge.net" rel="nofollow">http://breadcrumbemacs.sourceforge.net</a>.<p>I use it constantly everyday as I built it for myself. Judging from the download stat, not many people using it.
I'm not sure this is the "best feature" that no one uses, but it is one that I've always thought some crafty hacker could do something cool with.<p>BonjourFoxy[1] adds Bonjour support to Firefox and is predominately used to find websites advertised by printers, network cameras and so on, but it can also be used by other extensions to advertise and discover network services.<p>Currently the only extension I'm aware of that makes use of this is iGiro[2] which combined with an iPhone app let's you scan bill numbers straight into Firefox's text field.<p>Now admittedly I haven't got any killer P2P extensions in mind, but surely some other folks out there do —?<p>[1] <a href="http://www.bonjourfoxy.net/" rel="nofollow">http://www.bonjourfoxy.net/</a><p>[2] <a href="http://igiroapp.se/" rel="nofollow">http://igiroapp.se/</a>
I built a multiuser blog for a site with two million plus users (soundpedia.com). I worked hard to make it scalable; making it static, easy sync with new servers in the cluster, etc... Themeable too. It was pretty sweet for the time frame.<p>Turns out we had an audience whose blogs were composed of copying lyrics and changing colours. You can guess not many really used it.<p>Oh well. SP went out of business a while after so I never had a chance to do what the audience really wanted (a like button)
<a href="http://www.travelatlas.org/" rel="nofollow">http://www.travelatlas.org/</a><p>Based on the terribly original idea of writing about your travel experiences. A couple good features (imo):<p>- Inherent credibility. I figured if you ask someone you meet on the street where's a good place to eat or stay - being a complete stranger - the chance of them giving you good information is about the same as flipping a coin: 50-50. So everyone starts with 50% credibility. As you write reviews and people are in agreement (via like/don't like or comments) your credibility goes up (influenced of course by the credibility of the commentor).<p>- Built in messaging and chat. Added some "bots" to make it look like there were actually more users at one point. The idea was to implement a parser so that you could ask someone in chat: "where's a cheap place to stay in guam?" the app would parse the query and a bot could answer with the highest rated hostel in that location for example.<p>- Parse search terms with ability to search by category. So if you search for "a bed to crash on" while you're in manhattan, the app prompts if you'd like to search the "lodging" category.<p>Reason for failure:<p>- Other people did it better (tripadvisor, virtual tourist, etc)<p>- SHOCKWAVE (although this was 2005)
I called it WorkSearch: <a href="http://www.linkup.com/worksearch/" rel="nofollow">http://www.linkup.com/worksearch/</a><p>Basically it is a timer for jobseekers that runs in the background, to push them to spend a longer amount of time hunting for a job than they normally would (even if it isn't on our site). People have always known that the biggest difficulty with job-hunting is spending enough time on it, but it is really demonstrated by the this 2009 NYTimes interactive graphic: <a href="http://www.nytimes.com/interactive/2009/07/31/business/20080801-metrics-graphic.html" rel="nofollow">http://www.nytimes.com/interactive/2009/07/31/business/20080...</a><p>Despite some bloggers writing about it (<a href="http://helpmegethired.wordpress.com/2009/06/" rel="nofollow">http://helpmegethired.wordpress.com/2009/06/</a>), it never took off. I'd say the failure came from the execution. It needed to be dead-simple to turn on (w/o requiring an account), and always present yet unintrusive. Maybe the next version...
On <a href="http://www.ratemystudentrental.com" rel="nofollow">http://www.ratemystudentrental.com</a>, it was Watch Lists. If there was no available rental property that matched what you searched for, you could just hit a button to "create Watch List" and RMSR would notify you when a rental that met your criteria was added to the site or became available. Everyone thought this sounded like a "great idea" for the week or so it took me to build. Now 2 years later, exactly 2 people have ever used it.<p>Why did it go unused (I wouldn't say it totally failed)? The biggest reason is probably because no one wants to use your site as much as you want them to. If they couldn't find what they wanted on my site right away, why would they stick around when they could just go look elsewhere?<p>That's why I ended up building the "Around the Web" feature, where if a rental couldn't be found on RMSR, it will aggregate and map properties from other sites (around the web). This feature was mildly more successful.
<a href="http://wunbar.com" rel="nofollow">http://wunbar.com</a><p>Quick and simple to make, thought I could make some money on referral fees, but went nowhere.<p>Failure, I suppose: browsers can do the same thing, with a little setup; and, more importantly, people seem to have a preference to navigate to each of these sites and search them directly.
I built a tool called Statzen that I shutdown because it didn't get traction. It had a feature that would show you what topics/tags from your blog were getting read (via web and/or rss) and what sub-topics within that topic were getting read most.<p>I think I should have just released that one feature as an MVP.
Ooh, I've got a couple. I'll just use one for an example though:<p><a href="http://college.mychances.net/tools/college-choice-matrix.php" rel="nofollow">http://college.mychances.net/tools/college-choice-matrix.php</a><p>This tool is based on ~25,000 college applications and shows you the relative preference for each college. It's not based on direct matchups but instead based on an Elo point system, so that all colleges can be ranked along one dimension (essentially, revealed student preference).<p>It's rarely used because it's really hard to find on the site. It's also tough to figure out how to choose the schools that show up, instead of just looking at the defaults. It's also possible that college applicants care way less about this sort of thing than I, a processor and curator of their data, do.
The multiplayer mode of my Android game "Laska". There are just not enough active players to make it work.<p>But I like the feature because its based on AppEngine and XMPP, so it doesn't have any costs and will probably be available forever. (forever on the internet means sth. like 2-5 years)
Hmmm feature... not really.<p>I did have a website that pulled multiple RSS feeds (thousands) and then parsed each entry matching it to others that it had seen before. It could then show you what was popular at any moment on the web, with the ability to see what the article was linked to, and what they were linked to. It showed some interesting things like how a website was effected by something happening on the world stage.<p>The idea was you could find what was hot on the web at any one time. Nobody really showed any interest in it though.<p>Reason... I didnt want to pony up the money to put a live version on the web due to not knowing how the producers of the RSS feeds would fee. So there was never anything more then a demo concept.
<a href="http://BuyersVote.com" rel="nofollow">http://BuyersVote.com</a><p>Chicken/egg problem with reaching critical mass. My plan was/is to grow it organically with long tail keywords once Google starts indexing all the pages. But they haven't done that yet because the site doesn't have enough incoming links to warrant indexing 1000 pages (another chicken/egg problem :)
It's a tie for me.<p>1. The entire <a href="http://www.choip.me" rel="nofollow">http://www.choip.me</a> website, I built it because I was sick of seeing ads for twitter scams on other tweet longer sites, so I just threw together a Twitter + Disqus mashup on appengine to do the same thing, and provide a better way to take discussions "offline" on Twitter. Not bad for basically a weekend project I think.<p>2. The custom search portals on www.unscatter.com. Of course I'm still working on it, the interface for creating them is very rough. But the functionality is there. A couple examples of what can be done with it are the Devsearch - <a href="http://www.unscatter.com/search/unscatter/devsearch/" rel="nofollow">http://www.unscatter.com/search/unscatter/devsearch/</a> and Conservation Search - <a href="http://www.unscatter.com/search/unscatter/conservationsearch/" rel="nofollow">http://www.unscatter.com/search/unscatter/conservationsearch...</a>
I just built a site for the CrossFit gym that I go to that allows members to track the results, flag various milestones, make easy comparison to previous workouts and comment on other people's results.<p>They initially loved it, but when they found out it couldn't be integrated into their Wordpress blog, they dismissed the completed project. So I guess the reason for its failure was lack of input.<p>The project didn't take that long to create and was made to be more of an advertising tool for my company.<p><a href="http://crossfitter.ca" rel="nofollow">http://crossfitter.ca</a> (although you can't see that much unless you belong to a signed up gym)
<a href="http://bettersoftwareprocess.com/" rel="nofollow">http://bettersoftwareprocess.com/</a><p>I always wanted a site like this to discuss software process things without the hardcore evangelists using buzzwords. It didn't take too long to throw this up so it wasn't a massive waste but I actually thought it would start growing once I shared it with my developer/agile network. Not much interest, apparently. I'll probably have to kill it soon so it doesn't sit there mocking me.
I built a social networking site by myself with real time chat, messaging, photo sharing, music sharing, users can post wall messages, display their public timeline tweets, create friends list. I think the reason it didn't take off is the lack of money in advertisement. I tried Google adwords but the most I spent is 50 bucks and kind of short of it. The site is <a href="http://www.jamafriend.com" rel="nofollow">http://www.jamafriend.com</a>
I built an events search that aggregated data from several sources - upcoming, last.fm, etc. It didn't work out, mostly because even with multiple sources of events and a lot of data massaging, it was hard to get more than a page of results for any location / topic. At its best, it would just show a page of random results for a givne location, which was probably what it should have been in the first place (an aggregate calendar).
I wrote a jetpack app the allows you to toggle the Search header, Left Column, and Footer elements on/off in iGoogle.<p><a href="http://jetpackgallery.mozillalabs.com/jetpacks/55" rel="nofollow">http://jetpackgallery.mozillalabs.com/jetpacks/55</a><p>I'd say it didn't take off because it requires the user to have firefox, install jetpack, use igoogle, and be annoyed by all that wasted space etc :)<p>I wouldn't say it failed. I use it at work and home, and I love it :)
Pulling live ebay auctions to match with regular price comparison. It seems that the most effective way to improve conversions on a price comparison site seems to be just to put relevant links to merchant sites front and centre on the landing page, the majority of people coming from Google don't care what fancy value add features you have created.<p>Freelance work btw, I don't run a affiliate marketing site myself.
I've got a personal todo application that's aimed for lazy people. I've seriously tried a dozen of "smart todo", and no one were good for me. The one I built is special because it is so simple, flexible, and it does more than what a paper agenda can do.<p>I could publish it but I would need to modify lots of thing to make it intuitive for new users, write docs, and basically just to make it "public".
<a href="http://code.google.com/p/django-linkcheck/" rel="nofollow">http://code.google.com/p/django-linkcheck/</a><p>The take-up is probably limited by the fact that my documentation is erring towards the minimal.<p>That and the erroneous dependencies I forgot to remove until last week. :)
This one: <a href="http://www.karminator.com/" rel="nofollow">http://www.karminator.com/</a><p>I thought this would be a competitor for <a href="http://www.fmylife.com/" rel="nofollow">http://www.fmylife.com/</a><p>Not just any feature, the entire site is something that none uses
I built a Twitter client that filters out everything in your feed except for Tweets including urls or photos. It's pretty good way of just finding the interesting stuff in your Twitter feed.<p><a href="http://blurl.me" rel="nofollow">http://blurl.me</a>
Supported the entire 3GPP2-A10/A11 protocol for my product called Unsniff based on one customer's strong demand. One day my customer contact mysteriously disappeared and no one has used this feature ever since. That hurt.<p>Lesson learnt !
<a href="http://www.vidteq.com" rel="nofollow">http://www.vidteq.com</a> . We are a self funded startup. Usability is bad and performance issuses. How can we drag more people to use this application?
The site<p><a href="http://www.soundsabitlike.com" rel="nofollow">http://www.soundsabitlike.com</a><p>probably it's not visible enough on Google or it's just me who finds song similarities interesting :)
We spent months on an in depth analytics tool for our users.<p>Most of them don't have an idea what the data means. It's over their heads, so we missed the mark audience wise.
our app allows users to interact (create reminders via natural language) through email. I added in a feature/easter egg that lets the user start any email with @answer, followed by a question. Instead of reminder being added to our system, they are sent back the answer to the question. Uses natural language recognition
I wrote a jetpack app the allows you to toggle the Search header, Left Column, and Footer elements on/off in iGoogle.<p><a href="http://jetpackgallery.mozillalabs.com/jetpacks/55" rel="nofollow">http://jetpackgallery.mozillalabs.com/jetpacks/55</a><p>I'd say it didn't take off because it requires the user to have firefox, install jetpack, use igoogle, and be annoyed by all that wasted space etc :)<p>I wouldn't say it failed. I use it at work and home, and I love it :)