Ask HN: Here is my current batch of ideas. In exchange, I ask for your feedback.
In sharing these, all I ask is that you provide feedback on them, even in the form of "I hate 1,2, don't care about 3, 5, 6, and I may be interested in 4."
1. An alternate twitter landing page (e.g. username.twitteridea.com) that each user can heavily style and customize. I'm shocked that Twitter customization is relegated to simple background images and color. Tweets are pulled via client-side javascript, so themes are simple html/css/js. As awful as MySpace skinning was, people clearly wanted it. Viral potential is large here. Main problem, people don't read a lot of tweets by visiting a user's profile. Revenue model a bit iffy.
2. A simplified system for conducting beta tests. Match up companies (thinking micro ISVs or small webapps here -- this is not an enterprise play) that need exposure + a diverse tester pool with people who like being early adopters. Allow testers to earn reputations and allow companies to filter candidates through quality thresholds. Now testers are incentivized to do a better job, so they can get into more "exclusive" betas. Plenty of room to add in social elements here too. I think someone must be doing this but I've yet to Google the right set of keywords to find them. The differentiator here is that it's an offering for testers too, not just companies. Almost creating a marketplace where the currency is reputation.
3. A hosted solution for referrals (non e-commerce - this is not an affiliate system). I posted about this here once at http://news.ycombinator.com/item?id=1280478. Basic idea is to host a system/api that lets you build per-user referral codes, track those referrals, and then provide widgets to let you give users badges, show top referrers, things like that.
4. A guided daily diary. Inspired by Keen's Simple Diary. http://www.simplediary.com Pose simple questions to allow people to reflect on each day. Short and sweet, no feeling of obligation. Okay, that's just a rehash of the book itself. Now make it interesting by crowd-sourcing the guided questions or making them social (amongst your friends). Share entries etc. Favoriting answers. This needs more fleshing out. Big problems are: content until you solve crowd-sourcing, and how to monetize.
5. Private blogs/microsocial sites for families. Keep it fun and easy - think Tumblr. #1 use case is just sharing photos - email sucks for this, nor am I friending grandma on Facebook to share photos, and almost all blogging platforms I've seen besides Posterous do not offer a reasonable password protection system. But after photo sharing, I think family member activity streams could be fun, and you could integrate some location functionality so kids could "check in" easily without having to call/txt etc. There are a lot of ways it could go. I posted about this at http://news.ycombinator.com/item?id=1385447. The offerings I can find - myfamily.com, familysays.com, and famiva.com are truly wretched. Seems like a huge opportunity here.
6. Site monitoring that goes beyond the simple "is it up"/ping/L7 health check junk. Gomez and Alertsite et al. are fine solutions but stupidly expensive. Just stupidly expensive. Seems like there is a gap you can slide into here between those two ends of the spectrum. Real admins measure their page latencies not just uptime. (Note: just found out about http://www.watchmouse.com today, which sort of moots this... seems to be just what I envisioned at the price point that I imagined could succeed.)
32 comments
[ 10.6 ms ] story [ 75.0 ms ] thread#6 - Again, something I had started and had working for my own servers a few years ago, but never expanded or monetized. I don't think I have my old source code anymore though...
#4 I think vox.com does that. Whatever happened to vox btw? Is it just hanging in there, bringing in some ad revenue?
#5 I was thinking about this yesterday. Definitely a need. Photos + stories are huge, think family sitting around picture albums.
#6: yea, totally. I always thought site monitoring where you can send it events from your backend or js. Google analytics has events, but can't send it from backend (ie. signups, ...)
4 - not sure, never tried vox.
5 - no one seems to be directly attacking the vertical very well. i think "families" is a vertical, right? :) some things you can repurpose to get what you want, but the average person would never think to use it that way. the ones that are specifically built for families -- well they're either just awful, or seem like some kind of re-skinned groupware.
6 - you can do that in google analytics. it's a hack, but you can definitely post up metrics from the server side, if you just build up the urls that it needs. but anyway, this is definitely something where i was just thinking more about availability and performance monitoring.
I like 2 a lot as well.
Thanks for the feedback on #2. I'm starting to think I should make a go at it.
Edit: Not trying to criticize, i really like it.
1. Easy to implement, but I don't know if people would be interested in it. Another downside is that twitter can pull the rug under your feet at any time.
2. Interesting. I'm not sure if the benefits for testers are big enough.
3. I have no opinion on this; I don't like the referrals in general (except free "word of mouth") anyways.
4. nice but as with any social site it's hard to get the initial people and critical mass.
5. Hard to differentiate enough or see the real benefit when there are so many blogs/facebook/etc applications out there.
6. There are definitively lots of monitoring services of all kinds, prices (from zero) and features. It would be interesting to have one addressing a very particular niche one or at a particular feature set / price point as you said.
1 - totally agreed it's a Twitter feature waiting to be vacuumed up. but it's SO easy to implement. and I know how to make it go way viral. i think? :)
2 - why does anyone beta test anything? i think people just like feeling special, having special access, being early adopters, being influencers, etc. plus, i imagine the site as totally having badges/rewards built into it by design. and version.future - framework to allow companies to reward their testers (name in credits, direct payments, tip jar, whatever)
4 - agreed. maybe i could use #3 to solve this problem :)
5 - i know there's a market for this one, just from being a parent and knowing lots of parents. i don't think facebook is acceptable since there are weird issues with friending your parents etc, and grandmom is probably not even on the site. blog-wise, there are verrry few blogging platforms that have the right security controls.
6 - agreed. it would be easy enough to try something simple: 10 urls, 10 sampling nodes, 10 tests an hour, $10 a month - and just see what kind of interest is out there.
I like #2. I've been thinking about something similar (Feedback for Startups - http://www.nilkanth.com/2010/05/21/6-ideas-off-my-chest/)
Simple Diary (#4) looks cool, but I'm not sure if it can do more than a hobby.
Family network (#5) sounds interesting. Facebook, and other family networks, are just bloated.
2 - interesting. did anything come of you getting your ideas off your chest? :)
4 - agreed. it's probably a hobby or in the best case a one hit wonder.
5 - i think something really customized to the problem could be successful. everyone is forced to shoehorn social networks or blogs (e.g. CMSs) into the solution. i really picture something that's fundamentally different, while incorporating elements of each.
So I had this idea to create a distributed monitoring service that would use nodes (similar to the genome@home project) with actual residential internet connections to give websites a true idea of how fast their site was from different locations around the world. Node owners would just be regular people, and they would get a small percentage of the revenue from each test they ran, and the infrastructure would be designed such that a failing or slow test result would initiate verification tests from other nodes to insure that it wasn't just an issue due to the node's internet connection slowing down or failing. Gomez runs different test from their own servers around the country and sometimes we could tell what the issue was based on which nodes were down (i.e. the west coast server is the only one that's failing). So we'd have all these nodes running all over the world and we could tell people "well here's how quickly this page and all it's resources load from New York City, LA, and Russia..." and give them historical and real time metrics and notifications based on those response times.
I got as far as creating a client that worked relatively well and was ready to upload results for each test ran to a master server, plus it made requests to the master server to see what the next test it needed to run was. With all the other availability testing services out there I kind of moved on, but I've still got the source if someone is interested.
I would consider #5 as the weakest idea in the bunch. There are an enormous number of existing sites that offer the same premise. In fact, one could even use Facebook in the same way when the family restricts the friends to other family members.
#1 would be super simple and I predict very popular. Could build an ecosystem of template designers ala wordpress themes.
#2 All I can think of is mechanical turk.
#3 takes the work out of having to build all that yourself.
#4 I imagined this like a daily email service a la Groupon instead of a webapp.
#5 How does this not already exist?
#6 I'd like to steal this one but don't have the time
For example, we have checks to see if there are fan, drive, power supply, etc, failures within systems. We also monitor main power feeds, UPS, ventilation, etc on site infrastructure. You can also monitor/graph/check history of websites and their load times.
EDIT: you might also want to check out ganglia from the OS side. This doesn't allow notifications (as far as I know) like nagios but you can instantly see what your resources are being used on in larger machine clusters.