Ask HN: review my startup concept - web achievements

I think you are on the right track with the idea - but I see some problems with some of the specifics.

1) A lot of companies are paranoid and will want to run this inside their network/firewall. You should develop that option (look at GitHub firewall for example)

2) What incentive do platform providers have in pooling achievements / users - I don't see that happening. I would focus on a white box rewards platform - think SimpleGeo for rewards.

I wouldn't focus on having Webchiever being the destination for leader-boards ...etc because then you are competing with the people you are allegedly providing a service for. Become the plumbing, and if you do it well people will pay.

Cool idea.

Congrats! I had this idea about 8 months ago and knew that it was going to exist in some form on the web within 12 months. Now there are two services that are entering the space, so that is very cool.

However, my original idea seemed to be a combination of both webchiever and iactionable. And to be honest, I think both are necessary to make a successful service. Since founders of both companies are in this thread, I'll just lay out my thoughts here for you guys to incorporate as you see fit.

Fundamentally, achievements and points are a form of motivation/incentive that the designers of software/service can use to reward behavior they want the users to do. Now this has one really big assumption, that the users of the system actually care about the points and badges. For my high level thoughts around motivation check out my blog post (http://mysimplemindedworld.wordpress.com/2009/04/16/a-framew...)

So going by Bartle's classification, the people that are motivated by points for their own sake are the achievers. And for some sites, having a few achievers do a heck of a lot is enough. However, if your leaderboard only has 5 people on it, the competitive aspect isn't really there and so achievers aren't AS motivated, and you don't get the Socializers/Explorers.

So you want to have quite a few people part of your leaderboards to spur competition, set standards, and show off their stuff. However, for small sites that can be difficult.

The solution? Aggregate points/achievements across many sites/services. Follow the standard of Xbox Live and have the meta-game of overall Gamerscore be a big motivator. So while that one particular site may not have been that interesting to you, if you know that you can increase your overall NetScore by another 100 by contributing to it, well that's just the extra bit of motivation you need to do.

You then have a centralized identity that people can show off. You also then have a service that sites will want to use to motivate their users. Of course you have the standard chicken/egg problem, however there are enough sites that want to outsource this kind of point/badge system you'd probably have it cracked.

Here are the specifics for my idea... which I liked to call SocialScore.

There are two point systems at play. There is what i like to call the "internal" point system, and then the "meta" point system.

Internal points are applicable only for a specific site, and are used to power a site's leaderboards. So let's use twitter as an example. If they signed up to SocialScore they could create an event such as "get a follower" and that would be worth 100000 points. Writing a tweet they could make 5000 points. Following someone is 10000 points. Whatever, totally arbitrary and up to the site owner. It's only used for insite comparisons anyway. You then have a leaderboard widget system, and also show leaderboards on the central site.

You then have a badge system. Just like Xbox Live each site that signs up gets a fixed number of SocialScore points, and an upper limit of badges they can give out. Maybe you start with 100 SocialScore points you can give out (per user) and 20 badges. Of course there needs to be a verification/approval system in place to make sure people don't just register a ton of fake sites to get lots of badges. The integrity of the system is very important.

The interface/rule engine I concepted would consist of the following. A site master would register a series of events with the SocialScore system. So would register "Get a follower=1000 points" and the system would return an event id. And there could be an arbitrary number of events a site could register.

The site master would then create a set of requirements on which to award a particular badge. For example To Get Badge "Follow Master" a user needs to have event id XXXX happen 100 times. Or you could get a bit more complicated and be event id X happens 10 times and Y happens 5 times and Z happens 30 times. You get the idea.

Now on the programming end you have a very very basic API that just registers events. It's just one call "registerevent(userhash,eventid)" and the SocialScore system handles all the point calculations, badge awarding, etc.

The SocialScore site can then provide a bunch of widgets that client sites can embed such as leaderboards, profile-badges, etc. I'd even have the SocialScore system have a javascript library that would handle notifications, etc. But that's expansion stuff.

The business model here is great too... your first block of SocialScore points and badges are free, if you want to be able to award more SocialScore points and badges site masters have to pay for them. I disagree with the per-event payment model that Webchievements is pursuing since there's too much risk for the site. Also, if the aggregator site gets enough traction there are sponsorship and advertising opportunities. There's the obvious regular advertising, but then there's the cooler advertising where sites will want to be promoted to users of SocialScore.

So there's my original design laid out. It seems to be more ambitious than what both IActionable and Webchievements are going after... I don't know if that's a good or bad thing.

Just from my rudimentary math skills, I'm not sure the pay-per-achievement assignment would work. As an example, let's say I have 10,000 active users per month, and from their activity, they achieve 10 things per month. At one penny per, I would be shelling out $1000 per month for your service, $12,000 per year. Simply put, hell no. Perhaps, a better pricing plan would be in tiers with fixed costs, and notifications if one were nearing the limit of their chosen tier?

Also, for that $1000 alone, I could probably hire a developer to build an achievement system on my own server for my own sites. Just curious, what would the incentive be for me as a site owner to pay for your service as opposed to building my own?

You watched

http://g4tv.com/videos/44277/dice-2010-design-outside-the-bo...

, didn't you?

I've never tried adding achievements to my app...is it difficult to get right?

The services I like to outsource are the ones that are hard enough on their own...mail server (sendgrid), possibly hosting (heroku), recommendations (directededge). Some of them which are relatively easy, like comments with disqus, have been successful, but there is an added benefit here of letting users comment with the same account all blogs they visit.

Are achievements hard enough that users should offload them to you? Will they have to define various rules using some sort of interface you develop? I'm skeptical at first glance, but have been wrong before.

This reminds me of this project by Jeff Lindsay (progrium) : http://www.getachievements.com

I'm not sure it's still active though.

Problem - if a new game shows up, it doesn't provide a fresh new leaderboard for the achiever to climb.