Ask HN: Review Our Idea (Achievements for the web) | Hacker News Reader

Ask HN: Review Our Idea (Achievements for the web)

blog.iactionable.com

21 pointsryanelkins23 comments

Perhaps I am too much of a "normal" but the phrase "SaaS Achievement Engine" sounds like that specific type of tech-marketing-speak that will be unappealing to site owners because they will be too bashful to admit that they have no idea what it actually means. I have no such hangups.

Second comment is that I think incentive systems are highly highly tied - almost inextricably so - with the content and community around each individual site. In order for a reward system to work correctly, it has to be based on existing user behavior. For site owners who know this, they'll probably want to structure the system themselves. For site/app owners who don't, how will you sell them on the idea that incentive systems are important to begin with? I think the market you're dealing with right now is people who understand the importance of incentive systems but don't know or are too lazy/busy to make one themselves. If you believe this market is large enough, you should go for it.

The whole SAAS Achievement engine thing is a tricky bit. We've found that people that are familiar with achievements from games immediately get what we are talking about. Those that aren't struggle with the idea a bit. The post does explain what an achievement system is and we do realize that educating people will be a hurdle to overcome.

As far as how tied together things are we have definitely been thinking of offering some professional type services where we do alot of the work for people or some consulting work to help them best implement this system. Those who know what they want will have free reign to implement whatever kind of system they want as nothing is predefined.

I'm just beginning the process of adding badges to my app, but it wouldn't have even crossed my mind to use an outside service for it. I think in some form your idea has legs, for example as a WordPress plugin or other situation where people need simple plug-in achievements, but not as something to integrate in a case where the developers could do it themselves.

Sending you events just so I can query you right back to see if an achievement was unlocked? That seems like total overkill, I wouldn't even have the time to spare for the HTTP calls. And I wouldn't want to introduce a dependency. And all this for probably pretty trivial checks in most cases.

Yes, we addressed this in one section of the article. This is probably the one thing that gives me the biggest doubts. Many achievements can, like you said, be very trivial. Either way, it's alot of extra work that needs to be done constantly. Optimizing and implementing that is something that we don't expect everyone will really want to do themselves. Achievements can potentially be a very large system with alot of code, rules, conditions, etc to manage and that's the one thing that makes me think that it could require an entire separate system, much like a blog (Wordpress), forum (PhpBB), customer feedback (Get Satisfaction, User Voice), and so forth.

As far as the whole plugin thing - we've thought of that too and are exploring applications that we can create open source plugins for so someone can very easily plug our system in. For an example of how that might work, just try leaving a comment on our blog itself ;)

You need to sell this idea better. Right now it feels very much like a solution without a problem. You talk about stickiness briefly, but without specific and concrete use cases I'm not convinced. You might want to investigate thesixtyone. They use an achievement system quite like what you describe.

Ask them how they think it works, and whether their traffic would be the same if they removed the feature, etc. More data is great, because the argument doesn't catch my eye right now.

I've used a couple services with badges like that, but I've always ended up finding it stupid after not too long. You might end up accidentally alienating people who just want to use a service without feeling like it's a competition. Once I started associating a useful service with the time wastingness of a game, I started reevaluating whether I really wanted to use the service after all. There are games I far prefer playing, that I could just play instead if I really needed validation that badly.

If a feature is improving traffic numbers for the site owner, but not really improving the core usefulness of the application, I think it is a net loss.

This is something we are definitely thinking about. Part of the problem is that there aren't alot of application out there with a system like this so it's hard to measure the effect at the moment. Right now It's just sort of a guess. We'd definitely be interested in working with anyone that would like to test this with us to see what the effects are.

I've actually been thinking of doing something like this for a while. I think the market is there, but I don't think it's SaaS providers. I think you should start looking at web games and social games. Also take a look at Facebook games. There's a LOT of Facebook game developers out there. Granted, it's moving very much so from individuals to large companies, but you might still see some customers there.

As for implementation I'd definitely use a server side API. It would be nice to be able to send a large list of the user's data and receive back any possible achievements. For example if I had an RPG and the user had just killed a monster I would want to send you my health at the time of kill, what level I was, what level the monster was, how many of that monster I've killed, how many monsters total I've killed. That way I could get back multiple achievements if they were "Kill 10 Monsters" and also "Kill a Monster That is a Higher Level Than You" but only use one API call.

Anyway, I think it's a good idea. Try asking some web game developers.

I was thinking about a "Web Achievements" toolbar for Firefox etc. Where people would get awards for being rickrolled, spending time on websites, visiting certain websites for x times and so on. Maybe with a (adblockplus like) subscription thing that allows any site to provide achievements for itself.

It's not like I will ever pursue this idea, so if someone thinks this is a good idea, please use it!

Great idea! I would like to see this being applied to something that spans a whole lot of discrete sites, like Disqus for blog comments. Then allow badges for number of comments total, high interaction on a particular site, etc. It would be interesting to see your achievements follow you around the web.

We actually hope to address this part of the market with our orignal site which was http://www.KaBadge.com - Here we would display all the badges a particular user has earned across many sites.

We had also discussed creating various blog-engine plugins to automatically pull these badges and put them by your name when you comment on posts anywhere.

I would have liked a 2-3 sentences summary.. I don't want to be rude, but I stop reading in the middle of the first paragraph. I know you have worked really hard but try to sum it up in a clean and easy way. (And by the way, "SaaS Achievement Engine" tells me nothing.)

The problem is that it's not an entirely easy thing to sum up in a few words. What part of SAAS Achievment Engine doesn't make sense? I don't mean that in a rude way, I honestly want to make it better. I've changed it to "achievement engine service" and added a brief summary. I realize its alot to chew through so hopefully that makes it a little easier to determine if you want to keep reading or not.

Please let us know what you think! There are some questions and alternative ideas at the bottom of the article and we would really like everyone's input.

Edit: Here are the questions in case the article was too long for some :)

- What are your thoughts on the idea as proposed?

- Is this something you would consider adding in to your application at the right price?

- What is the right price?

- Are you interested in the full on achievement engine or would you prefer an API that just awards badges to users directly?

- Would you benefit from a Rule Engine which could handle complex rules with many conditions? Meaning, rules more complex than just ‘Comment Event occured 5 times, Award Badge’.

- In order to prevent people from spoofing API calls we may expect sites to call our API functions from server side code instead of via a javascript/client library. Is this asking too much? Are you willing to integrate our API at this level?

It's not important enough a problem to solve. If I want badges and stuff, I can very easily program that myself, without the headache of integrating with you guys. I don't think this will work. On the other hand, email delivery is hard and important, so I will pay a company to do that. Find a harder problem to solve for people.

I appreciate your honest feedback. What did you think of some of the alternatives we have been playing around with?

If you were to make your own achievement system, would you share those badges with a repository? This is basically where we started - http://news.ycombinator.com/item?id=795952

When I was at Yahoo Brickhouse, we developed something very similar, which half-launched but appears dead:

http://bravo.yahoo.com/teaser/

I never felt comfortable with the project primarily because I didn't feel I understood the market at all.

Yeah the market is something we don't understand to well either which is a large reason for this post. I remember seeing someone mention this site before - any information you DO have about the market would be appreciated. This is sort of a brand new area and I'm not too sure how to go about researching it.

you could do it with client-side calls as long as the server-side can sign the data (ie use an hmac). that still means a server-side api, but it doesn't have to make the calls.

for this to work i think you have to make it really easy to use and with minimal impact on the developer. would you handle the generation of badges (this seems like it would be a timesaver - you provide a badge editor that has basic options like colour, etc)? and serve them?

there's some tension between tying awards to a particular site and making them global. initially i assumed people would want their badges displayed as an integral part of their site. but you could also imagine some "global" badges that you provide that people can earn from various sites, and which people might want to include in blogs, personal profiles, etc...

We plan to provide a JavaScript library and CSS file that can be used to both get and display the badges. People are free to use this themselves or create their own way of displaying badges. That makes that part of it very easy on the developer while still giving them the flexibility to do something completely different if they'd like..

Didn't read all of it, but I think it might be similar to an idea I once submitted to YC. Also reminds me of http://sf0.org/

I think it can work. But I also tire of game mechanics by now - I don't want external influences on my choices.

Achievement systems have 2 issues: a. users attempt to fixate on achievements more than the software they're supposed to be using b. users cannot achieve certain things, realize that its hopeless, and lose interest in the system all together and possibly the software

Yes, there are some interesting side-effects when you attempt to motivate people to do very specific things. But, I don't think people leave the site or software when certain badges are too hard.

I think http://www.StackOverflow.com is a good example of where badges only add to the experience and don't get in the way.

The right set of achievements paced correctly can counter your second point. The first is quite valid, though.

Reply on news.ycombinator.com