Ask HN: review my startup concept - web achievements

by bonquesha99 • 16 years ago

12 19 16 years ago

yourabi2• 16 years ago

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.

bonquesha996• 16 years ago

Something standalone like github:fi would be pretty cool, thanks I'll keep that in mind.

I was planning on allowing providers to pull achievement data back out like you're suggesting (unless I misunderstood you) but I'd also like to have a centralized listing of achievements as well. It allows users to discover sites/service that their friends are using, which would drive new targeted traffic to those providers. I think users would be more willing to check out sites/services if they see that they're friends are using it. If a user is already using a particular service and sees that a friend of theirs just got an achievement, their natural urge to be competitive could drive them to check it out and get the achievement themselves.

It could also be useful for helping users decide things in marginal situations. Imagine a user is hungry but not yet motivated enough to do anything about it - seeing an achievement from a friend where they got 10 points for ordering a pizza online could help push that user in the direction of completing that achievement, and in doing so driving revenue for that provider.

tonystubblebine2• 16 years ago

I definitely agree with you that achievement systems could be usefully applied to the real world. I built a point system recently to manage my own life and have now become addicted to flossing. I submitted for a demo account but would love to talk to you more generally. tony@tonystubblebine.com

Also, are you Sean Huber of InquryApp?

vyrotek3• 16 years ago

Thats interesting that you say 'become the plumbing'. Thats exactly what everyone told us when we first launched http://www.KaBadge.com - Why do I want to give you my badges?

That's why we ended up creating another product called http://www.IActionable.com where we provide a rule engine that constantly evaluates actions performed by users and awards the badges, points, whatever for you.

I was actually just reading someone's comment this morning about how they created an Achievement system for the workplace. Can't seem to find the post it was on though.

Keep up the good work guys. I'm happy to see other sites enter this space.

Here are the posts we made for our two services. Perhaps the feedback provided there can help.

(IActionable) - http://news.ycombinator.com/item?id=1126780

(KaBadge) - http://news.ycombinator.com/item?id=795952

bonquesha996• 16 years ago

You definitely seem to be on a similar track! IActionable is interesting, I hadn't really considered trying to handle the rule aspect of assigning achievements, though I can see for some sites how it certainly would be valuable. Have you had any luck getting traction?

barmstrong• 16 years ago

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.

vyrotek3• 16 years ago

As far as I understand, the service provided will be a respository of badges. Your site will have to determine what the badges are and when to award them to users. Then you'll let Webchiever know about it so they can store a copy. This way users have one place to show off all their badges.

In regards to how difficult it is. The repository concept itself isn't hard, but the idea is if you can get enough sites to use you then I imagine other sites will see the site as a great advertising opportunity and will want to award badges just to get visibility on the site.

The rule engine is a different story. (Of course I'm very bias) A 3rd party rule engine is useful if you plan to constantly add badges and/or change the conditions which award a badge. Otherwise you need a developer to essentially rewrite your badge awarding queries or code every time you want to make a change. Or write your own flexible rule engine with a management UI. :)

bonquesha996• 16 years ago

You've pretty much got it, especially the 2nd paragraph. And your right they would need a developer, or at least some familiarity with the inner workings of their own site in order to insert the code needed to award an achievement to a user. They can also manually assign achievements to users, or generate achievement claiming codes and send those out to users as they see fit.

franck• 16 years ago

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

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

Sidnicious• 16 years ago

You watched

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

, didn't you?

bonquesha996• 16 years ago

Absolutely.

OmarIsmail• 16 years ago

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.

johnrob• 16 years ago

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

bonquesha996• 16 years ago

Could you clarify what you meant by this? When a new game is created, the developers of the game will be responsible for defining the achievements in webchiever.

hkuo• 16 years ago

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?

vyrotek3• 16 years ago

Just so its clear, with this idea you would have to build you own achievement system regardless. You're just giving the site a copy of it.

Just curious, what would the incentive be for me as a site owner to pay for your service as opposed to building my own?

It really depends on the complexity of the 'rules' which award the badges and how often you update the requirements or add new badges. If your badges are going be like "post a photo!" then you really don't need one. But, if you have a very high traffic site and a badge that is awarded if X happens 100 times & Y happens 200 times OR Z happens 3000 times... then you have to consider how you will manage these requirements. And also how much extra load you're adding to your database every time a user performs one of those actions.

Ask HN: review my startup concept - web achievements

Related Stories