Tell HN: Missing service – StackOverflow, but with monetary rewards | Hacker News Reader

Tell HN: Missing service – StackOverflow, but with monetary rewards

21 pointsbenevol53 comments

There is a market for technical questions that few people _can_ and/or _want_ to answer or work on.

When you need to do research for hours (or even days), you are of course willing to spend a few bucks to get your solution quickly, and get on with your project.

- So you would describe your technical problem, as usual.

- You would then set your monetary reward for the accepted answer.

- You could increase your monetary reward if answers are lacking or coming in too slowly (the list of unanswered questions could be sorted by "reward").

- To make your start easy, you could import your StackOverflow "karma" (points) into your new profile.

View on HN

That was Experts Exchange model before StackOverflow ate their lunch primarily because it was free [1] as devs would rather answer questions for no money (i.e. to help people & build rep) than little money which would turn the motivation from the feel good act of helping people to whether it would be worth it to spend their time answering questions, which for most devs it wouldn't be.

[1] https://www.joelonsoftware.com/2018/04/06/the-stack-overflow...

Could it be that times have changed, evolved? Or that the SO folks don't know or understand everything?

I've had the same idea. Operationally, what's needed is speed.

I do not care to learn some obscure, poorly documented API by trial and error. If I could write a failing test or three in a playground then post it in an appropriate chat window and have two or three people claim it at once, with the winner earning 50% of the prize, I would utilize this with some frequency.

The problem with an SO-style solution is that I don't get confirmation that anyone is working on it. So if my project is under a stiff deadline then I have to keep trying myself.

That said, yes, someone please build this. There is a market for people like me that have wasted days on ffmpeg just to get X to do a Y without a Z.

Or overnight. If you could post a simple problem to fix and its done by the next morning that would be awesome. Would play to strengths of Asia time zone too.

> There is a market for people like me that have wasted days

Days? I know I must have wasted weeks, in total.

And not to mention how this can pull you down emotionally/mentally, when you just don't find the solution to your problem. Having someone pull you out of a deep hole is worth money.

> Days? I know I must have wasted weeks, in total.

I meant for a single bad interface!

In total I shudder to think about all the wasted time.

How much money tho? It's unlikely anyone will put in more than a seventh of 1.75k for an app like this.

Putting in a chat window turns it into crowd sourced peer programming. You could get people who do nothing all day as a profession but this. Like github copilot but with real developers on the other end.

re "someone please build this", Literally, I can do it for 1.75k if someone could gather the funds.

Joel, one of the founders of StackOverflow, has a blog entry about why this won’t work:

https://www.joelonsoftware.com/2006/08/09/the-econ-101-manag...

It is worth the read

People get crazy enough about virtual internet points. Introducing real money adds a lot of problems you'd have to solve. Especially how you determine whether an answer meets the requirements and should be paid out or not.

Managing taxes, withholding, reporting, and other legal requirements in many thousands of individual jurisdictions would be a challenge as well.

Doubtful that it would work. Many, if not most questions on stack overflow are poorly written. Good luck moderating that with payment incentives to people who provide an answer. Or if someone does provide an answer and the question asked simply says it doesn’t work in good faith or bad faith. Or if someone provides a bad viable answer before you get around to submitting an excellent answer and you lose the reward. Or if a bunch of non English speaking folks start spamming the forum with low quality links to random pieces of documentation.

The power to determine who gets and does not get the money is going to be a very difficult moderation challenge.

And on top of it; the money pool is likely to be very small imo, because there’s already stack overflow which is free.

There were a few attempts at this in the past, even before StackOverflow existed and they always failed. I remember listening to a podcast which dissected the success story of StackOverflow and how it improved on its predecessors. One point I remember taking away was that all predecessors that used money to attract answers were less effective than the counterparts that solely relied on non-monetary rewards such as upvotes, karma, a simple "thank you" and other kinds of social approval.

Sorry for not being able to back this with the original sources.

Lots of people here are saying "it won't work", and that _has_ been historically true. However, there does seem to be a gap in the market for ways in which to pay open source software developers.

I currently have "3.6m people reached" on my Stack Overflow page. I have no idea if this is low or high, but if it was converted to YouTube views or Spotify plays it would pay me about $150,000-$300,000.

Instinctively I feel like there probably could and should be some kind of way to reward people for contributing to (something like) Stack Overflow.

SO profiles are, like Github profiles, a fantastic resource to put on a CV. In a vacuum, it doesn't give much, but it's a great skill amplifier. It also gives visibility from SO itself.

Afraid looking for social / FOSS participation further disadvantages minorities and those less well off. Since they have less free time they may be less likely to have active profiles. Hell just becoming a parent has caused my SO activity to drop off.

And some employers may not want GH activity made public for concern about leaking working hours or time zone, such as in defense.

Not having a SO profile is a non issue, for most of my career I didn’t have one. I still boycott LinkedIn as well out of principle. Neither of this has prevented me from getting clients and jobs.

I’m sure some wacko hiring departments demand a super active SO profile. They’re irrelevant to the larger discussion tbh.

If you believe hiring departments shouldn't use such signals then how can they be a "fantastic resource to put on a CV" and "skill amplifier"?

Demanding as a requirement is not the same thing as using as a signal.

So it's OK to use signals that push already disadvantaged groups further down the list of candidates?

This is a complex question that doesn't have a yes/no answer, and you're wording it as if you're allergic to nuance and want to bait me into a flamewar. That is the literal definition of "flamebait".

Having a particular skill, such as "knowing javascript", is also a signal that "pushes already disadvantaged groups further down the list of candidates". So, you tell me.

You're not the first person to think about these issues. This stuff is no different than people publishing their portfolio on artstation, or even on their personal website. And on SO specifically, it's also very similar to having on your CV a list of places you volunteered at: Unpaid jobs to show off your skills.

Sorry if I've come across as hostile. Just trying to understand how you reconsile what looks like a discrepancy to me.

As an individual I can see listing anything that gives me an edge. As a hiring manager I have to be careful to be fair and feel an obligation to avoid signals likely to filter already disadvantaged groups. So for a manager these considerations do come down to a binary: will this signal be used?

> Having a particular skill, such as "knowing javascript", is also a signal

I'm not advocating removing requirements undeniably needed for a role.

SO, volunteering, and unpaid work aren't generally needed for roles I hire for. Therefore as signals they're less useful, and may tend to bias in favor of those who already have other advantages in life.

> You're not the first person to think about these issues.

Did I suggest I thought I was? And what does it matter if others have considered these things? All of us with hiring power each need to learn what is appropriate, or not, in order to make the best decisions.

I’ve used Fiverr for quick solution to WordPress site. I created three test environments and hired three freelancers to address my problem. The results were underwhelming, unusable, and a bailout.

This was for something that should have been straight forward for PHP developer.

But for something you want to implement that needs research? Yikes. All the worst behaviors of SO, but with a race to the bottom. Most minimum answer all the time, ignoring your edge cases, telling you that all of your boundaries are ridiculous and their way is the only way.

Is Microsoft answers community have paid experts? I dunno, but they’re always polite in that same way. And they don’t answer the question you asked, but politely. It can take several attempts to get them to recognize the issue. Only then you might get a reasonable answer. Why is that if not the economic pressure to answer as many questions as possible as quickly as possible?

Look at paid help from a different direction—customer service or technical support for anything. You know they’re paid agents with training, paid by the company they work for. They always have a bias.

If you pay for SO like questions, where is the bias? Can you figure it out? How can a user contextualize answers?

It seems like conditions for low quality answers.

We had that. It was called Experts Exchange. SO ate its lunch.

Yeah, the basic idea of Internet supporting a market for knowledge like that was great.

I don't know their business played out, but lots of dotcoms seemed to focus more on appearance of growth, and IPO cashouts, and the earlier obvious globally-connected-world utopia ideas didn't always do so well.

Experts Exchange's biggest legacy might've been to gain acceptance of "-" in domain names.

I asked a question a little while ago at a small stackexchange satellite site. It got a few upvotes but no responses. I used as much bounty as I could to feature it. It still got no responses. I was a bit bummed they didn't refund the reputation, but whatever.

I am quite confident that a working professional could give me an excellent answer in about 2 minutes, and I have to say subjectively the question is pretty interesting (not a "how do I <x>" or "do my homework" question). It is unusually hard to Google, Wikipedia doesn't have the right information, and there's only one major enthusiast forum that might have users that know the thing. I'd be willing to pay a token amount, maybe $20, to get the answer. Not that I think $20 is worth getting out of bed for, but given you were already using a Sunday afternoon to answer a few questions, not a bad minutely wage.

I think my main concern is that such a site is likely to become essentially micro-projects, more "do this for me" than "answer my question". And I think people who are paying would be even angrier than usual if their question gets moderated away for doing this.

> "do this for me" VS "answer my question"

The boundary is probably not always easy to define. And so I would ask myself - would there even really be a need for such a boundary? Why not simply increase the reward, if your question/problem is bigger than most?

Well, in this case, you're describing Fiverr with a slightly better user interface for the hirer and an incomplete contract problem (because you expect someone to do the work before you award the payment, and also reserve the right not to award the payment). This is fine but it's basically something we already have.

What I've needed in the past is more like "I need someone well-versed in this technology to answer me some questions for one hour" without actually finding a consultant or the billing hassle.

Basically just like joining the project's IRC channel and hoping that someone has time - but with a guaranteed best effort and quick response.

How is this different? It's usually not a single question where I am able to find an answer, or the community is actually quick and correct enough with an answer - it's more like "hey we used $tech for $thing and we have a feeling we did something wrong because $weirdproblem" and just scree nsharing actual closed source code would be a better explanation than trying to find a minimal repro case.

In a past company we actually tried to find someone like this for a few specific OpenStack things but had no luck. But it wasn't enough to book a consultant for an extended period of time.

Expert Networks can do this.

https://www.codementor.io/ sort of fits your requirements.

But behavioral economics says:

> Sometimes asking someone to do something for nothing is more powerful than paying them. [1]

Also, StackOverflow incurs zero monetary cost. There is a stark difference in people's behavior towards zero cost, even if the price difference is only 1 cent. [2][3]

[1]: https://danariely.com/why-bankers-would-rather-work-for-000-...

[2]: https://web.mit.edu/ariely/www/MIT/Papers/zero.pdf

[3]: https://youtu.be/WS1bwMdgmKc

The guy from CommerceHero launched his own private slack workspace for eComm that does essentially this. It's a regular slack hangout with a bunch of different relevant channels like platform specific ones, but also it has a bunch of really smart people. Kalen, the founder is trying to solve this problem by 1. setting an expectation one of your peers may be asking you a question and to try to give it your best and 2. going out and asking people. He hit me up and said, I know you're doing xyz, can you help this guy in #operations and see if you can help?

Seems to be working so far. I had a business process question around credit card processing that did not have a straightforward answer and a guy came out of the woodwork with useful things to consider, what impacts the decision one way or another and what he's seen in the past.

Yes, this is consulting but commodified and made more efficient on an open market. The legacy "consulting" business model probably includes lots of costs and inefficiency priced in such as time wasted getting set up with a new team, time wasted in meetings, cost of sales, insurance etc.

How much value can really be added though in the question-answer format that you can't already get for free on StackOverflow?

The answer probably lands somewhere on a sliding scale between "quick answer to an easy coding or api question" at 0 and "team of programmers speccing out and working on a project with delivery by a deadline" at 100.

StackOverflow currently sits at 0 to 10. Gig sites like fiverr at 50, and legacy consulting at 80-90. Maybe the value here is at 25 or 30 on that scale?

> How much value can really be added though in the question-answer format that you can't already get for free on StackOverflow?

Or a tier higher - for SO points. Whenever a question is not answered, throwing a few hundred/thousand points at a bounty gives you so much attention, I don't think you'd be able to get a better answer for money.

If average pay for a dev is x per hour, I wonder if you could work out how much one SO point is worth based on how much time invested to get the bounty

Uh, isn’t that called consulting?

Also I don’t think this is a Tell HN…

Where else could you get such quick/easy "consulting" on such a simple question-basis?

Wasn't that the goal of experts-exchange?

> There is a market for technical questions that few people _can_ and/or _want_ to answer or work on.

You think there is a market.

Joel Spolsky and Jeff Atwood discussed this on their podcast back in the early days about making Stack Overflow. They believed that having monetary rewards would create perverse incentives and the psychology of involving money would discourage people from participating - people would balk at $n/answer/vote/whatever thinking it's not worth their time, but would happily spend their free time answering questions for free.

What will happen is that people wil compete on speed (first to get accepted gets the money don't they?) and the quality of the answers will be comparable to first tier of outsourced tech support.

This is exactly what happened to ChaCha.

I miss the earlier days of the internet where the spirit was more about sharing and less about selling, advertising and competing for "likes".

Agreed.

When sharing knowledge was to actually inform someone out of pure altruism.

When gaming was about gaming and not money/sponsors.

When social communities (forums) actually gathered people out of pure desire to share common interests.

Those days are long gone, but we get to tell the stories.

Yes, let's introduce commerce and spam to a system in which people already expend a lot of effort against spammers seeking fake internet points.

Please, no.

Let's stick with a site that is mostly populated by people internally motivated to provide good answers, rather than a site that is mostly populated by people externally motivated by money. The answers will, on average, be poorer, guaranteed.

How would you handle a scenario where an answer is posted (which could arguably be acceptable) but the poster doesn't acknowledge it as accepted?

Or realizes a few hours later that for a complex enough problem it was not a sufficient answer?

Is there an arbitrage service? Because StackOverflow mods already go crazy with power. Add money to the mix ...

arbitrage? Do you mean escrow? And yes, problem is easily solved with a 3rd-party deciding how much to award.

What is the incentive to accept answers?

But more importantly, there's the Jon-Skeet-already-has-a-job problem.

And the handling-payments-is-a-PITA problem.

Ultimately if someone has the money to make an expert answer worth their while, they can hire a consultant.

And if an expert wants to make their living answering questions, they can open a consulting firm.

Good luck.

It would be neat if you could get this to work.

But I can't help but think anyone who launched such a service would end up spending most of their time adjudicating disagreements about whether there should be a payout when a poorly stated question gets a low-effort unhelpful answer.

This is not remotely a new idea. Also it seems like literally any short term contraxting site already fits this mold and more.

I used Codementor in the past and I loved it, you pay by the minute and you get expert help in my experience.

That a perfect fit for a Web3 service.

https://gitcoin.co/bounties/funder

Reply on news.ycombinator.com