Why don't bounties ever work on this site?

Question

I've offered a lot of bounties for key questions I think deserve better answers. I've seen other people do the same, or of course offer bounties on their own questions they're hoping to get more info on. Almost universally, the result is silence.

In September, there was an overhaul of the bounty system, which gives the ability to offer reasons and comments on a particular bounty. That seems good, but I don't see any evidence that it's improved the situation, at least not on this site.

The few cases where a bounty is awarded seem generally to be highly-active users who might have provided a good answer had the question been brought to their attention in another way.

Is there a way we can make this better?

Maybe the reputation reward isn't enough.

Can we offer cake? Should we focus a contest around bounties? More badges?

Maybe bounty questions don't get enough publicity.

Is there a site-wide way to advertise them? Something more we can do off-site?

Something else?

Is it a network-wide problem, or something specific here?

Answer 1

Having collected 9 bounties and raced for a few more, I wouldn't say they don't ever work. There have certainly been some questions I would have dismissed into "don't know, don't care" bin, but the bounty and lack of answers has switched on curiosity - maybe it's a hard question worth tackling?

I suspect that the smallness of the site means that easy questions will be answered in short time anyway and hard questions are, well, hard (usually requiring really thinking about the details and/or researching some materials). Getting multiple people throwing in good answers not duplicating each other is a slim chance.

Fetching an awesome answer for an old question with voted answers might be beyond powers of bounty; you're probably better off crafting out the answer on your own. If there are already several good answers with many votes and maybe even an accepted answer, it's unlikely that a new answer would ever get enough attention to float to the top, even if it really is awesome. Why dedicate time to write something awesome only few people will ever scroll to?

But is it really so that bounties result in total silence? According to StackExchange Data Explorer, 74% of bounties set on Photo.SE have been eventually awarded (38 out of 51). These odds sound like 3 out of 4, better than most gambling options :) Stack Overflow seems to be doing even slightly worse, with only 70% of bounties getting eventually awarded.

Certainly some things could be improved, especially visibility. I have sometimes edited a bounty question (or its existing answers) primarily to bring it back to the front page. It would be useful if one randomly picked "featured" question would actually be featured on top of other questions every time.

Answer 2

I think as a first step we should have a new silver and gold badges. There are currently a handful of bronze badges which encourage dipping one's toes in the bounty system, but nothing beyond that. I suggest:

Mercenary [silver]: answered at least 20 questions with pending bounties, received at least 5 bounties

Bounty Hunter [gold]: answered at least 100 questions with pending bounties and received at least 20.

Or similar. Actual values to be tweaked as appropriate. They could also hinge on amount of rep rewarded, not the count, or in addition to the count.

Answer 3

I think there are a few things going on here.

The Talent Pool

Photo exchange is a fairly small site, so there are a limited number of people that will visit it. Of those, a reasonable number of them will predominantly either use the site as a reference resource, or ask questions. I’d imagine there is a minority actually answering questions. Getting people to answer their first question would seem to be the main initial hurdle (if somebody answers + gets a positive response they’re far more likely to answer again). Bounties may incentivise some people to post their first answer, but I think it’s unlikely as initially 10 reputation is a reasonable reward for an answer and the bounty isn’t guaranteed.

The Types of Bounty Offered

Looking at some of your bounties (such as the ones you’ve got running at the moment) you’re essentially asking for better answers for questions that already have answers (in some cases accepted ones). I think the bounty changes will help a bit here, in that at least you can state clearly what you are looking for from the bounty. However, it’s going to be harder to get people to contribute to questions that already have existing answers because they have to be convinced that they have something worthwhile to contribute and the likelihood of this reduces with each existing answer, so again this reduces the pool of people that are likely to offer new answers.

Amount of attention

It’s true, that higher rep active users may have answered questions without the bounty if their attention had been drawn to the question. Editing a post moves it to the front page, so to an extent there is already a way to get the attention of active users. If you have a blog / know a community where other users can answer the question you can ask them to do so by providing a link to the question (this may even gain you an Announcer/Booster/Publicist badge). Your badge idea may also contribute to higher rep users answering some bounty questions (although I would imagine this would be a fairly short lived advantage before those users collected the new badges).

Expectations

Bounties aren’t guaranteed to produce a satisfactory answer (they’re a gamble). However, they should result in a higher likelihood of the question being view and hence answered (although as I’ve said to get a new answer people have to believe they can answer it). With that in mind, perhaps a better measure of the success of the bounty system is to look at the relative page views, once the bounty is added. This will show if the page is being viewed more + the bounty is working as an incentive.

Whilst the benefit of bounties on small sites may be less (questions aren’t rushed off the front page most of the time) I do believe they offer some benefit and have been awarded for some good answers, so removing them doesn’t seem worthwhile.

Answer 4

Another option:

Extend the time that bounties are available until the site gets bigger.

This would increase the amount of exposure a "featured" question gets; maybe the right person will come along after all.

Answer 5

I'm in the habit of checking bounty questions, but honestly most of the time the bounty is really asking for some kind of plug-in-the-variables formula type of answer when the real answer is actually pretty vague (and has to be). We're a technical lot, but photography is only tangentially a technical process.

I've seen questions about phychophysiological reactions to tone, line and colour, and all we really have to go on is anecdote (possibly because there is more cultural influence than we're willing to believe possible). Asking for a more in-depth answer to a question that the best current research can only answer with "we don't know" is a pointless waste of everybody's time.

The bounties don't interest me at all. I'll answer a question if the question is interesting and I have enough knowledge or experience to answer it -- red tags and rep are not incentives for me (nor, for that matter, are badges -- the only one I've "gone for" was the Fanatic, and I only bothered with that because I noticed that I had 90-odd days consecutive and figured "what the heck").

Am I an oddball? Well, okay, sure, but I mean in the context of StackExchange? I'm here to share a love for something magical I've enjoyed for some forty years, not to "win". Maybe gaming works as a concept, but it's not my thang.

Answer 6

On the network-wide question, several people suggest that bounties don't work on small sites. I'm not sure I agree, but if it's true, several things can be done:

A) Turn off bounties completely until the site reaches a threshold in traffic/users.

or

B) Decrease the cost of bounties to question-askers until such a threshold is reached.