Hacker News new | past | comments | ask | show | jobs | submit login
How We Make Money at Stack Overflow (stackoverflow.blog)
578 points by ocoster on Nov 15, 2016 | hide | past | web | favorite | 217 comments

> We are in the process of sending out an email now announcing Developer Story to the larger community in the hopes that we can help some percentage of developers. If that email is rejected, we’ll unsubscribe you. If you’re unresponsive to emails we send, we’ll unsubscribe you. If you mark it as spam, we’ll unsubscribe you.

How great it would be if every site would subscribe to such high standards. Bravo, SO.

Most sites do this already, at least they should if they want to make sure more of their emails don't end up in junk folders. Deliverability is really important to both businesses and email service providers, so a lot of the time, bounce unsubscribes and spam unsubscribes are baked in and easing off subscribers who don't open or click emails is a best practice for deliverability purposes.

Yes it's pretty much common best practice

Mail providers offer sending to your webhooks or block them directly on their side (so you dont have to worry about it)

Many do that probably, Indian firms don't for sure, I always block them, spam them and complain via twitter and after weeks of follow up do they remove my email id from their spam emails.

> If you mark it as spam, we’ll unsubscribe you.

How would they know?

In addition to the header info mentioned in other comments, email providers let you subscribe to email Feed Back Loops. You register for FBLs (man it's a neverending process since there's a never ending pile of new email providers), and when someone marks something as spam, the email provider (gmail/outlook/etc) will send an email to a registered address, you just listen for emails on that address. I wrote a custom handler to do this, as we already had a system in place to process incoming email from users (respond to comment email instead of replying on web form etc). I then unsubscribe the address in question.

Not all providers offer FBLs, in that case my internet provider gets an abuse notice which is forwarded along to me. Not a ton of activity there on a 500k person mailing list, the FBLs are pretty effective.

Also, since we have the email processing system in place, it's easy to detect bounces and auto unsubscribe people. I immediately unsubscribe on a hard bounce, and wait for 5 soft bounces before doing the same thing.

You can include a header in your email message where, if a user marks a message as spam, Gmail / Hotmail / whatever will let you know.

Why do they do that? Seems great for this use-case but terrible information to give bad actors.

When I've usually clicked the report spam and unsubscribe button after reporting spam in gmail it usually immediately opens a tab with the unsubscribe URL.

I've since stopped doing it since the last thing I want to do is open up some random page belonging to the person who spammed me.

That's strange. I thought the point of that button was to overcome the exact problem you just described.

Tried it again with a different message and they said they'd try to unsubscribe me this time instead of opening a link.

Did try again yesterday with a different spam message and had it open a tab though; Zapier was spamming me with some newsletter so I clicked "Report Spam and Unsubcribe" yesterday and it opened up this in a new tab: https://zapier.com/app/settings/email?payload=REMOVED

Not really sure how it determines that -- I'll probably avoid clicking it just out of precaution and just report spam going forward though.

Using a lot of the privacy and security measures suggested in https://www.privacytools.io/ I sometimes do a gmail search using 'is:unread in:all unsubscribe' and click on most of the links I can find.

In some cases, Gmail automatically detects the unsubscribe link. If you click Gmail's unsubscribe link, it may give you a pop-up or it may automatically write an email to unsubscribe@example.com. The pop-up usually leads to a form that you have to fill to 'update your newsletter preferences'.

What information does this reveal besides your email address, which they clearly already have?

It's a feedback signal showing success at passing automated spam filters.

It confirms that the address is valid, and most likely attended by a real person.

But there are gazillion email leaks. Go scrape them all with hashed passwords. Active emails have no value.

Leaks may have emails that are defunct, belong to dormant accounts nobody reads, throwaway accounts like mailinator.com, honeypots, etc. This confirms the account belongs to a real human actively reading the email.

You're right that they don't have value but there's still plenty of idiots out there buying and selling useless lists and while they're worthless they have a cost to us as recipients since we have to mark them as spam and delete them.

It reveals that their email was detected as spam by an intelligent piece of meat instead of a shard of silicon.

Email lists with verified activity have a higher success rate and sell for more.

Many times when you hit the spam button, gmail ask if you want to "only mark as spam" or "mark as spam & unsubscribe".

Would make sense to prevent themselves to be marked as spam in the future.

idk, I'm trying to find a way a bad actor could take advantage of that information. Kind of helpful for spammers to get a sense of what does and does not get marked, but they can just see how many people clicked the link.

In the past (10+ years ago), clicking the Unsubscribe link in a spam email was a sure way to get your email address sold to every spammer out there.

Most of those bad actors seem to have been weeded out, but some still might try it.

In the comments, they explain that Sendgrid gets notified of spam marks, and Sendgrid gives SO a webhook whenever someone marks as spam. So they then unsubscribe that user.

They probably assume if you make as spam, you will also be unresponsive to the email. If you are responsive they can say that overrides the marking as spam action.

See the List-Unsubscribe mail header.

Sorry, if this is off-topic, you hit a nerve.

The absolute antibehavior of what you descrcribe I've recently experienced with Air Asia.

A few month ago they started to spam me with a newsletter I never requested. Last time I flew them was in the beginning of 2015 and I never subscribed for their crap.

The unsubscribe link on the mail does not work and after a significant amount of digging I found a way to unsubscribe by sending mail to unsubscribe@fly.airasia.com.

Only that mail to this address is never delivered.

In essence there seems to be no way to unsubscribe from their list.

I mention that because it's infuriating that "reputable" companies think they can spam away with gusto without even the possibility to unsubscribe from their shit.

Here's a hint for you Air Asia: You are not reputable! You are on the same level as pushers of fake knock-off Viagra, purveyors of fake shlock Rolexes, and Nigerian scammers! And you will never, ever get an additional cent from me.


I think you're pretty on topic. Airlines...

I have had a very similar with Air France who I flew with once, through a "hacker" booking of a flight provided by Delta but through Air France's site to save hundreds of dollars over Delta's.

They actually subscribed me to four newsletters: "Flying Blue", "Hello from KLM US", "Air France USA", AND "iFly KLM Magazine United States". From those I've received ~75 messages over the past year.

Their unsubscribe links do not actually work. They prompt you to log in, but I can't log in to my account because I don't recall ever making one (and I'm diligent about storing passwords in a manager). Attempting to reset my password leads to a screen that says I'm locked out of my account for 3 unsuccessful attempts, and the only possible option is to call them. It's so broken. </rant>

A perfect CAN-SPAM complaint opportunity.

I'm not sure if they're required to conform to it being outside of the US. A quick Google search sounds like it might be based on whether the mail server is in the US or not.

Despite never having Comcast nor living in an area Comcast serves, I constantly get email from them. To unsubscribe I must login to my (nonexistent) account.

This may actually be against the CAN SPAM Act of 2003, which requires that non-transactional unsolicited commercial mail provide a means of being removed from the mailing database through the same means that the mail is sent.

IANAL, so I don't know if logging into an account would count or not. Personally, I see it as a hindrance to unsubscribing and by my reading of the CAN SPAM Act, it is illegal, but.. not a lawyer.


From https://www.ftc.gov/tips-advice/business-center/guidance/can... (which isn't the Act itself, but an FTC compliance guide):

    Honor opt-out requests promptly. Any opt-out mechanism you offer must be able to process opt-out requests for at least 30 days after you send your message. You must honor a recipient’s opt-out request within 10 business days. You can’t charge a fee, require the recipient to give you any personally identifying information beyond an email address, or make the recipient take any step other than sending a reply email or visiting a single page on an Internet website as a condition for honoring an opt-out request. Once people have told you they don’t want to receive more messages from you, you can’t sell or transfer their email addresses, even in the form of a mailing list. The only exception is that you may transfer the addresses to a company you’ve hired to help you comply with the CAN-SPAM Act.

Maybe not sending an email in the first place?

If their product is so great, they don't need to spam anybody, specially developers, to go look at it.

I really like email as a discovery channel for new products personally. Is there something else or another format you prefer?

Probably going to get downvoted for the unpopular opinion but community curated news sites ala. Reddit/HN are way better at this than private communication channels like emails.

I spent a big part of the past year in email marketing for another company... It is a huge job for any organization, I'm glad that SO has such high standards. Making the product as good as possible, over making interaction the goal in and of itself without thinking of the need for that interaction.

In lecture 6 of "How to start a startup", entitled "Growth", Alex Schultz touches on this point specifically in a hypothetical concern people might have of being too spammy. His point was that one should worry about the marginal user who doesn't get notifications, not about the power users who are grown ups and can set filters. Granted, SO is no longer a "startup" but growth is still relevant to it nonetheless and it is interesting to note that the question of emails is far from being over and done with.

How about opt-in first?

If SO here I'd like to extend my thanks to. I might be interested in "offers" but it'd be nice to detect when I don't give a sh*t about it and back off.

SO is doing it right. But the community will very likely not care.

Great answer to a related FAQ in a post[0] linked from this article, about why !SO doesn't care about ad blockers:

> The truth is: we don’t care if our users use ad blockers on Stack Overflow. More accurately: we hope that they won’t, but we understand that some people just don’t like ads. Our belief is that if someone doesn’t like them, and they won’t click on them, any impressions served to them will only annoy them-- plus, serving ads to people who won’t click on them harms campaign performance.

[0] https://stackoverflow.blog/2016/02/why-stack-overflow-doesnt...!

And a great explanation in the piece of how SO works to serve only quality ads.

Good enough, in fact, that I'm whitelisting Stack Overflow from my adblocker. I don't do it often, I don't even do it for sites I like; the reality is that anyone using an automated ad network is at significant risk of serving malware, and I don't approve even if I'm protected. I don't visit Forbes at all anymore, ever, because they broke their users basic faith in "see our ads, and we won't destroy your machine".

But a promise to serve relevant, low-impact ads, chosen by actual humans? That's something I'm willing to support.

Did you only do it today? 8 months after the announcement, because someone happened to mention it in a comments thread? Doesn't speak very much to that approach's effectiveness. Especially since you've got a strong opinion on it.

The fun thing is that I think Reddit once used both types of ads. I don't think they do that anymore though.

Just because an ad network is automated doesn't mean you can serve malware through ads. Just because an ad network uses manual bidding doesn't mean it's malware free.

Both true, but I think my point is clear?

Specifically, this isn't about manual bidding, it's about manual review, and I trust the SO people to keep out likely malware candidates about as much as I trust any site.

And yes, there are ad networks that haven't served Can't is a word I might reserve for text ads, but there are other automated networks that haven't at least.

But I can't imagine its confusing why I might prefer manual review (malware policy: "Oh god, we missed it, never again") to automated distribution (malware policy: "Thanks for letting us know, we'll take that one piece of malware down in the next day or so.")

There is also a nice feature: if you have enough reputation there (I'm among the top 1%) you will have an ad-free experience.

I'm guessing at your level of being among the top 1%, you'll likely not need to actively seek work; instead great employers will seek you out hence their conscious decision not to annoy you by showing you ads.

But then you also have trackers [0].

[0] http://meta.stackexchange.com/questions/237062/does-stack-ex...

StackOverflow is an extremely valuable resource for all programmers. I've not seen the option anywhere, but a yearly donation would definitely be something I'd do for a substantial amount (substantial to me as a student, so it's not the moon but not a beer either).

My following suggestion is a very unpopular one, but it is this:

If you are a student, use the money you would have donated to StackOverflow to purchase a book or a tool that will help you to learn better and become an expert.

If you are an expert, contribute your time to StackOverflow (instead of money).

Money is easy to get compared to content. What helps StackOverflow grow is quality content.

I am also interested how they approach the subject of money:

> How much money we make is a direct proxy for how much we are helping our users. We focus on maximizing how we help users in order to make a great business. Very few companies have done this successfully, and we take great pride in the fact that we’re one of them.

Personally, I would love to see how this works out and I wish them well. Hopefully they can prove this model works well. Users need this model to work well for other businesses to adopt.

>>book or a tool that will help you to learn better and become

I've tried that before, many times. Unfortunately when you are trying to do real work those books are many times useless. Example, I kept trying to build a newtonian physics simulator so I got a Statics and Dynamics mechanical engineering book. The book was a text book for Mechanical engineering students and yet the examples they teach you are for toy problems. I could not generalize them to real world scenarios. To actually build the physics simulator I had to scour the internet until I found something that actually explained what I needed in an easy to understand manner [1]. And that still was not enough. I still had to use more online sources to find out all the other pieces, like the separating axes theorem for collision detection.

So no, I would not rely on books anymore, at least not exclusively since they often do not provide a full picture. One last thing, books seem to be getting progressively worse in my experience. All of this is anecdotal of course.


As an aside, buying a print reference book frequently feels counterproductive.

Even if I'm going to get a formal reference text, instead of online resources, I probably want a digital copy. Otherwise, I find myself going "I wish I could do text search for this thing..." and wondering why I didn't get the version that actually allows that. When I read for pleasure I enjoy the physicality of books; when I read for technical reference I demand the interface of a computer.

For books without ebooks from the manufacturer but for which I can find an illicit PDF or djvu, I usually buy the print one and read the digital version.

I have the same experience with books. I don't buy non-fiction for learning purposes* because it never worked. I tried with an IPv6 book which I read a few chapters of; I tried with another book or two that I don't even remember; I was gifted two books, one about making games and the other about programming paradigms, neither of which I ever read more than a few pages of... and yet I'm perfectly proficient by just playing around in spare time and looking up stuff I need and stuff I run into.

* I do buy non-fiction for amusement and casual interests though, like Predictably Irrational.

I've contributed a fair amount to Stack Overflow, but it's actually pretty hard in the niches I care about. Not because the questions are hard, but because people answer them so quickly: it's hard to find a question I can answer that hasn't been addressed well. That's a great problem for Stack Overflow to have, of course, but I do wonder if there's some easy way to monitor specific tags without having to constantly refresh the page...

There is, if you use an RSS reader. There are feeds associated with each tag. Try clicking any tag, and you should see a link where you can get an RSS feed of all the new questions that come in for that tag.

It's been there for a long time. ;-)

Well if there is no work left to be done, so be it I guess? I don't think you should feel bad about that :)

> Money is easy to get compared to content. What helps StackOverflow grow is quality content.

I do, >10k rep on the security.stackexchange.com site. I just think it's an amazing platform that I'd like to support with hosting and development costs.

> I do, >10k rep on the security.stackexchange.com site

That's amazing! Thank you for your contributions. I meant to imply that donating money is usually the least fruitful way to contribute, in my experience. It should be your last resort; definitely not the first one.

Example: One way SO can really gain from your security expertise is for you to think about how anonymous answers can be supported. That will be far more valuable than $$$.

I posted about this here: https://news.ycombinator.com/item?id=12961447

I would love to read your opinions about this crazy idea!

I have to agree. This happens on forums as well. I've Google searched something before and found people discussing my problem on some random forum. Maybe I know something that might help, but its a hassle to create an account to post.

I bet that happens on Stack Overflow too. People who'll never bother to create an account find a page about a problem through Google and they have something valuable they'd add it if didn't prompt them to sign up first.

I wonder how much trouble it'd actually cause spam wise.

I suspect there would be a lot of challenges with it but perhaps letting high rep users validate anonymous contributions would help with the spam. Reveal them only to authenticated users and you remove the incentive for SEO spam.

Contributing is wonderful, but those internet points (reputation) screws it up. An edit would give you 2 points if you have a rep < 2000, and contributes towards badges anyways. People would make pointless edits to a post. See http://stackoverflow.com/posts/21023161/revisions for example. There is no change between the original answer and the latest revision after 3 edits.

That example is all on Emil Vikström, who unilaterally accepted a bad edit from Akshat (with minor corrections of his own) after two reviewers had already voted to reject. That's two edits. The third edit was undoing the bad change they accepted. The undo was also contentious, with 2 votes to reject and 3 to accept.

But, Akshat has only 18 points. Even if you're lucky enough to have worthless suggested edits accepted, it's not a very good way to earn points.

Excellent insight. I am hopeful SE will figure out a way to stop/reduce this arbitrage opportunity.

We're humbled hearing from folks like you - your willingness to give back is what makes it all work. Keep your money. But here's how you CAN help, right now: Sign up if you haven't, create a developer story, and when you ARE job-hunting, try Jobs. We I think you'll find it to be the most developer-friendly offering on the planet.

Just an anecdote from my experience — I am really enjoying the new Stack Overflow Developer Story. Mine is at [1] for example. While I do appreciate jobs, I'm a full-time independent contractor and not looking for a traditional job right now (though I have seen quality offers through SO Jobs).

That said, Developer Story / CV has been instrumental for me and has become my canonical resume. I'm a huge fan.

If I could make one feature request... it would be a way to indicate conference attendance. It's something I give extra credibility to, for both employers as sponsors / presenters and employees as attendees / speakers / volunteers, as not everyone is willing to put in the effort to give back. [You can see on my CV that I've tried to incorporate this as an "other" at the bottom, but it feels a little out of place.]

[1]: http://stackoverflow.com/cv/taylor

I'm a dev in the Developer Story/CV team and we appreciate the love <3 The conference attendance item idea is great, discussing it internally with our team right now. Thank you!

Can vouch for it. I found my current job through SO Careers, and now that we're trying to hire more people, we're finding a good number of our better candidates coming through our SO Careers posting.

I would like to suggest a feature request...

I would love to use developer stories but more importantly, would like to integrate my dev story with my website (as this would also keep the "recommended reading" section up to date as well). If there was some way to do this (we can discuss options) without using an iframe, I would LOVE to do so.

Just to mention...I want this on my website & stackoverflow. Updating recommended reading is of a lower priority than having it on my website in the first place.

It is probably better that stack overflow can support itself. Donations would put them at the mercy of others whims. A donation model doesn't make sense when a company can sustain itself while still preserving its core values.

> Donations would put them at the mercy of others whims.

Yes, which is why I assume they didn't even bother putting up a donate button. The thing that convinced me to post anyway is that it's probably very clear to a lot of people that it is definitely worth supporting. Much more than the Archive project or the Tor project, this is something I know people use so much every day and get tremendous value from. That's why it might just work in this case, similar to Wikipedia.

It's also not exclusive, they can totally put up donations and try to attract some attention for it by posting it to their own ad system, and at the same time continue with all the current efforts.

I consider the time that I give to the community answering questions to the best way to help. While I have no idea how to value that in dollars, it's something I put an effort into regularly.

You can also buy a beer for folks who actually wrote an answer you used if you ever meet them IRL. I bet direct revenue from users would be negligible by the other revenue streams Stack Overflow Inc pursues.

"a yearly donation would definitely be something I'd do for a substantial amount"

There's a word for this, it's called 'subscription' : )

They provide a service, you pay for it.

It's how most of the world works.

I think the underlying issue is price discrimination: how to get the large numbers of those unwilling/unable to pay some value, while providing value to the higher order users as well.

It's a paradox that people will pay more for a coffee than they will something that can help them substantially :)

> There's a word for this, it's called 'subscription' : )

I avoided that word because it's the wrong word.

A subscription implies I get something that normal users don't get. I don't want that. I hate services like Dropbox where the free users get benefits for installing it on their Android or from referrals (shitty stuff) and live off of the backs of paying users. Yet here I think it's much more akin to something like open source software, which is for the general good and is worth supporting for those who are able to afford it.

Edit: There is no need to downvote the person, though. To me he seems to be contributing to the conversation.

"A subscription implies I get something that normal users don't get."

Fair enough - but there are very, very few value creating engines that can feasibly exist in such a context.

Wikipedia is one.

Stackoverflow probably could not.

Someone mentioned it creates 'billions' in value, well, that's probably not too far off.

There's nothing wrong with people paying for things.

But to your point - it would be nice to find a way to do intelligent price discrimination. Students, hobbyists get it free, other entities can pay. etc..

I don't think that would work for Stackoverflow either because there is an inherent 'open-ness' that it needs ... that said, again, there's nothing wrong with paying for things.

Stackoverflow is the prime example for how a company can generate tremenduous value without making that much money.

It's difficult to estimate its contribution to the world economy quantitatively, but I'd definitely say billions, many many billions.

How do you know they don't make that much money? I didn't see that anywhere in the blog post.

I agree that they make an enormous contribution to the world economy. Joel really did knock it out of the park -- remember expertsexchange? His goal was to get rid of that and it worked.

Joel, eh? Think you might be forgetting someone in there, somewhere? ;)

This is what I love about HN, occasionally the author of the software being discussed will randomly show up in the comment thread

Correct me if I'm wrong, but I always got the impression from the podcasts that while you both designed it, it was more you actually programming it. I could be mis-remembering after all this time!

Didn't you quit your job and program it while Joel was still mainly focused on Fog Creek day-to-day?

Yes, it was Joel's famous "servant leadership". We brainstormed ideas once a week on weekly calls (which became the podcast) and he trusted me (and Geoff and Jarrod, the first two people at the company after the founders) to build Stack Overflow as we saw fit.

Community feedback was a huge part of that feedback cycle once the beta was live.

You may be gone, but not forgotten, never that, Jeff :)

> Joel really did knock it out of the park -- remember expertsexchange? His goal was to get rid of that and it worked.

You left out the other co-founder: codinghorror [1]!


Joel [and Jeff Atwood] really did knock it out of the park -- remember expertsexchange? [Their] goal was to get rid of that and it worked.

[1] https://news.ycombinator.com/user?id=codinghorror

Or as Joel likes to call it 'that hyphenated site'

Good point.

A thought provoking example I had pointed out to me recently is the GPS system, which was built and is maintained by the US government but is a free public service. The amount of public value from it is astronomical.

i think they are like a bank. Value is inserted by people asking and answering questions.

Another good article linked from this one: https://stackoverflow.blog/2016/02/why-stack-overflow-doesnt.... Stack Overflow is IMO the site to look to if you want to do ads correctly. So many big sites have incredibly trashy ads, SO shows that they care to not subject their users to that junk.

I'm actually surprised ads work at all on StackOverflow. I would have guessed that most of their visitors are blocking ads.

Most of the ads are served to non-logged-in visitors who come to specific question pages through a web search.

SO actually shows you fewer ads the more reputation points you have.

What's the flow there for a user? They're researching an answer ... do they ignore the issue and run off and buy something? I sometime have poor focus but when you're problem solving it seems the least likely time to distract yourself with shopping.

I did once follow a Google ad about 10 years ago, but in my defence I was just surfing cool tech and the image looked interesting. Of course I didn't buy anything, who lets advertisers tell them what to buy?

I digress.

> They're researching an answer ... do they ignore the issue and run off and buy something?

Yes, that's an entirely credible idea. To use the first ad I found as an example, maybe I'm debugging an annoying issue with migrating my MongoDB cluster and I notice an ad for a managed MongoDB service and realize I'd rather pay them to deal with this.

> Of course I didn't buy anything, who lets advertisers tell them what to buy?

I do! There are millions of useful services out there and ads are an effective discovery tool.

Of course, we know you are special. But consider that other people may not be as strong willed and resilient to the advertisers' ways.

I think I would expect logged-in users to be more likely to have whitelisted the site than random Google visitors.

But there are a lot more of the random Google visitors than active participants. SO is effectively a machine for getting smart people to create specific, targeted searchable content to run ads against.

Stack Overflow is highly targeted tech content with a variety of detailed, popular and niche posts.

I suppose that they get high quality ads (i.e. pay lot more than the average ad + highly relevant so people click it).

If you turn off your ad blocker for SO, you get to see targeted job postings in your geographic and expertise area.

This is the way ads should be done. The only reason I run a blocker is because my tolerance for absolute bullshit ran out. No, there is no "one weird trick" to make hundreds of dollars a day and get rid of wrinkles at the same time. Just no.

I've seen the quality of advertising plunge to depths I never imagined in the last five years. I yearn for the days of "Punch the Monkey" level dribble rather than the pure poison we see today.

How is it that with more tracking, more compute power, more data than ever the ads are so appallingly stupid?

At this point, I really don't understand why people that know about ad-blockers don't use them. Do they also watch commercials when they are playing back a show they recorded on their DVR?

I honestly want ads. I want them to be of things I could at least theoretically want to purchase.

I've never seen an ad for a conference I should probably go to, or of a service that looks interesting, or of a product I want. Not in the last five years unless I've expressly visited their site and got tagged by Google, in which case all I see is ads for that.

Instead I'm barraged with bullshit news stories, weight loss products that are total garbage, wrinkle removal creams of dubious origin, and "suggested links" which never, ever deliver on what they promise.

The whole advertising industry is in shambles.

I paid a good amount of money for the jobs listings because it is Stack Overflow.

The sales person was very persistent and available while the negotiations were happening to use their service (yes, you can get discounts). As soon as I paid, she disappeared and was never to be heard from again. No follow up at all.

I also got zero good jobs references for an extremely profitable tech business in the heart of downtown financial district in San Francisco. All the people who applied were basically spam and that wasn't any better than just posting on LinkedIn or HN for free.

Not really a complaint, just one small data point to consider.

Perhaps it's time they consider paying Moderators for their time and effort. It doesn't need to be anything large, even a small stipend will suffice.

Moderators are the real lifeblood of StackExchange. They spend countless hours ensuring their respective Sub Exchange hums along. Yes, it's a volunteer "job", but it's a job none-the-less - especially since once you become a Moderator, it's more-or-less expected you have certain responsibilities and obligations. Given StackExchange makes a pretty healthy profit off the efforts of the "volunteers", and given StackExchange is a for profit company, it seems fitting they should have to pay for the labor that creates their profit.

Currently, StackExchange kicks back and rakes in money via various avenues, but then doesn't have to pay staff to actually manage the growing plethora of Sub Exchanges.

We all give companies grief for underpaying or mis-classifying employees, but somehow everyone overlooks StackExchange.

I believe it an accurate statement to say, StackExchange would not be a thing today, if it were not for the innumerable hours worked for free by a growing army of "volunteers".

I mean, sure, but I'm not sure why you say '"volunteers"' - they are exactly volunteers. Yes, extracting labour for free from volunteers can be the sign of something shady, but with SE I don't really see that evil intent.

Moderators not being staff means that the community has more say in how the sites function, which is a thing people want. It's hard - communities want to police themselves, not be dictated unto by a controlling body. As soon as you start paying the moderators, they are under your sway.

I also don't think it'd be easy to create a fair remuneration system that didn't incentivise the wrong thing, or create pressure on moderators that they don't currently have.

Overzealous moderators are the worst part about Stack Overflow.

Are you a moderator?

> Are you a moderator?

No, and have never been. I have no dog in that race - but as a long time Stack Exchange user, I witness these people's hard work every day.

> Every new email, except for transactional emails like “Forgot your password?”, sent through this system will have three links at the bottom: a one-click unsubscribe, a direct link to manage all categories, and feedback.

It's terrific that they're doing this. The first two links are what the CAN-SPAM Act requires. The "feedback" link is not a result of complying with CAN-SPAM, but is a good idea for any online business.


I like Stack Overflow, but I seriously don't know how all these answers end up there. I think about 80% of all my questions are answered sooner or later by myself. And some of them are quite massively upvoted too afterwards, so they weren't always niche problems.

A reminder that you don't need to use the latest or fanciest tech stack to achieve success in business. In the case of SO, boring wins, it's proven and seems like it allows them to focus on their product and to create new value for users.

what's their stack?

If curious, I did a post on that a while back. I'm settling into a new house and will pick this series back up soon.


Web Server: IIS, Database: SQL Server, Web framework (backend): ASP.NET/MVC. Additionally: Redis, Elasticsearch, HAProxy (all on Linux). jQuery on the front end (with some TypeScript) and LESS -> CSS.

TLDR: They make money through ads, job postings and their enterprise product. The author believes that they can make money and serve users at the same time.

In my opinion, they do a pretty good job of balancing the needs of users and advertisers.

Re: balancing the needs of advertisers -

We looked at advertising with them about a year ago and their options and response were both underwhelming. It wasn't possible to geo-target, it wasn't possible to advertise on most relevant sites (e.g. SoftwareRecs), and their upfront budget requirements weren't compatible with how we normally do things, which is to run a small test campaign to see what sort of traffic an outlet can drive and then expand if it goes well. On top of that, the response from their sales reps was curt and the last message in conversation, once it was clear that we aren't going to drop 50K, remained plain unanswered. So, YMMV.

Hi there! I'm Steve Feldman and I manage the Ad Ops team at Stack Overflow. I'd like to make a few leaps based on your comment and offer a little insight our business. I'm very interested in learning more about your experience.

- Based on your targeting concerns where we failed to adequately meet your request (geo, site, budget), I'm going to assume you tried to purchase advertising through our self-serve option. We offer self-serve specifically for advertisers who want to work with budgets below our minimum of $5,000 per month. In doing so, we hope to allow clients who still want to target our users an option while preventing our salespeople from burning out or being unable to provide an excellent advertising experience due to an excessive client load. Please remember: the long-term success of the Ad Sales team is reliant on the success of our clients and their ability to interact with Stack Overflow users in the smartest, most mutually-beneficial way possible. That takes time.

- Targeting via our self-serve option is limited to global run-of-site. This means ALL tags and ALL countries. Any advertiser looking to reach specifics types of users-- mobile, web dev, cloud-- will see better results working with a sales rep and spending that $5,000 minimum, as impression volumes will be too low to make a real, measurable impact through self-serve. We see greater success for self-serve clients with (a) products that benefit ALL developers and/or (b) products related to some of our larger tags on the site, front-end tags for instance, as the largest volume of impressions go to the largest tags when targeting run-of-site.

- Self-serve is currently limited to sites that have a large impression volume, ie: Stack Overflow and Super User. We don't currently offer any advertising on many Stack Exchange sites, and SoftwareRecs is one of them, unfortunately.

- If you can share the interaction with our team with me, I'm happy to look into it and make steps to speak with the entire team to ensure we're doing everything we can to have your experience with us be a positive one. If you're so inclined, send an email to adops@stackoverflow.com or to reach me directly, steve@stackoverflow.com.

I've managed larger advertising campaigns which would easily meet your $5k/m minimum but agree with the GP that your ad structure is a non-starter for me.

Whenever I've looked at introducing a new advertising channel, I always start with a test run of a few hundred dollars to see if it's worth investing more. I'm not going to drop $5k on an ad without knowing if I'll get any results at all.

It seems like you only offer two options for someone like me:

1. Run a test campaign without targeting and therefore get suboptimal results. This likely just lost you the sale of a bigger campaign (where I might get good results with targeting).

2. Risk spending $5,000 on a channel which doesn't give effective results. This isn't a risk I'd usually be willing to take.

Maybe you're targeting brands who are big enough to experiment with $5k buys or who don't test channels. Either way, I'm really not sure what the value prop is on the self-serve option.

Sounds like "if you're not spending the minimum then we don't want you ... unless you're a business that spend much more"? Surely the only businesses that would advertise without tag or geolocation specifying would be massive brands?

Are you shy about mentioning that you only want large advertisers; it's not possible you don't realise the ordering only makes sense for such companies, is it?

Or did I miss something?

Steve, to comment on just one aspect -

Geo-targeting is a must unless one's trying to build a "brand awareness" and need to spray everyone in sight.

In particular, in a context of trying to actually sell some professional/work software it's a norm and a common sense to target North America and Western Europe first and foremost. Conversely, and putting it bluntly, serving such ads to India is like feeding money into a shredder. So geo-targeting is absolutely essential.

I ask a lot of question, and thus get banned from asking more...

I wish they would have a "luxury" option for like 49.99 a year where students could ask like unlimited questions.

Is this crazy?

> students

Your question has been asked and answered, probably several times. You're getting banned because you refuse to learn about and use the search function.

Flooding SO with beginner questions lowers the usefulness for everyone.

There is no hard limit on the number of questions you can ask. If you are getting downvoted or banned you are probably asking low quality questions that are frustrating for the volunteers that answer questions on stackoverflow.

This is not a problem that can be solved by throwing money at it.

Those bans (which occur for users who post many questions while receiving few positive votes) are for the sake of the community of contributors to the site. Allowing them to be bypassed for a fee would risk violating the community trust that's emphasized in this article.

Almost certainly the quality, not the quantity, of your questions was responsible for your ban.

See http://stackoverflow.com/help/how-to-ask

After many years of being part of the Stack Exchange community I have come to the conclusion that folks will do whatever they can to avoid actually reading that page.

I've seen folks post questions on meta about how to get out of the ban (used to be a very regular occurrence, a bunch a day). I've seen folks post on other stack exchange sites asking how to get out. People even get emailed for this. All to ask questions which are answered perfectly well by the page they were linked to in the ban.

I've seen this so often that I can still remember the goo.gl URL that the old FAQ used to be behind (http://goo.gl/C1Kwu).

You may be interested in: https://www.codementor.io/

I have personally never used it and am unaffiliated with it, but a friend suggested it and has found it useful. He explained it to me as a dedicated, personal Stack Overflow.

how come there's no https://www.contradementor.io/ ? where am i to get an opposing point of view?

There is one. ;o)>

Not crazy but out of scope. If you need a mentor get one but SO is not for mentoring. On the other hand even 49.99 per month is too cheap for a service like this.

You could try offering that money to a TA for a limited number of email assists or phone calls. They're usually both pretty strapped for cash and people who enjoy helping.

What does TA mean in this context? Technical Architect?

There is a cost to other users in time spent sifting through questions, so it'd probably need to be a student only area. In a business environment this is something people pay for, but not $50 at a time.

How many questions does it take before you are banned?

There is no limit - the only issue is if your questions are marked as bad. If you ask questions that are answerable and work on StackOveflow, you can ask as many as you like.

If they started banning users for asking questions or even for asking poor (read extremely poor) questions, then the overall traffic would reduce significantly.

I wish. We've been doing this for 5-6 years now, and it hasn't. Quite honestly, if we could get rid of the worst 20% of questions instantly, life would be a lot better for the remaining 80%... But, identifying those 20% when often SO is their only option is... Brutally hard. I've seen the same person try to post variations on the same question dozens of times before finally getting around the quality checks... Only to have it sink like a rock because of course it was still unintelligible after all that.

Correct. This question - HOW MUCH MONEY does Stack Overflow make per year - was asked in their Meta site. Granted it's a privately held company but there is atleast 1 answer that quotes "credible sources" and pegs it at atleast 21 MILLION USD in revenues (for 2013).

Source: http://meta.stackexchange.com/a/214284/245495

According to Hoovers, their revenue is $11.33M. Source http://www.hoovers.com/company-information/company-search.ht...

How do they employ almost 300 people on $11 mil / year?

Also, why are these posts getting downvoted? The title of this article is "How We Make Money at Stack Overflow".

Venture capital. I highly doubt they're even close to profitable with revenue per employee of $36k.

I'm so happy we have stack overflow.

Do you guys remember experts exchange.com?

That site made me wish I could set up a "exclude from search results forever" on Google.

Chrome only, has be installed on each PC, not on mobile. Bleh.

I installed a firefox plugin to specifically remove those results from Google.

It was pretty good and usually useful IME until they got greedy and hid the content from all but subscribers.

It made me quite angry when they locked up my content. I switched to posting solutions on my blog then. Now I tend to stick stuff on Stack rather than blog it.

So is it cash flow positive? I hope so, because it's hands down the most useful site for programmers. I even got the gold medal for using it every day for a long time.

They make over 10 million dollars a year.

Between 300 people cited elsewhere that's only $30k. Presumably they pay for hosting, facilities. Doesn't sound like it leaves a lot of pre-tax pay unless they're using primarily overseas workers in low-cost countries?

None of their locations seem to be that cheap, if you look at the remote workers. Not Silicon Valley expensive, but not working for peanuts either.

$300k, surely?

      10,000,000 ÷ 300 
    = 1,000,000 ÷ 30
    = 100,000 ÷ 3
    ~ 33,333
(commas for grouping)

Or 1e07/3e02 = 1e05/3.


I have tweeted about SO several minutes ago before I discovered this topic on HN: https://twitter.com/otarch/status/798647799411675136

StackOverflow is so powerful resource on the Internet, alongside with Wikipedia and Internet Archive.

I wish I could donate two ways: 1) Yearly or monthly (subscription) donations, like on Wikipedia 2) I would be extremely happy to donate authors of some great, life-saving answers. I feel so excited when I find the right answer after searching a solution for hours.

Just shut up and take my money!

SO intentionally does not want to integrate payment of any kind for answerers because of Motivation Crowding Theory[1], the idea that intrinsic motivations like altruism can be crowded out by extrinsic motivations like profit, which is problematic because intrinsic motivation is generally stronger and effective for longer than extrinsic motivation.

[1] https://en.wikipedia.org/wiki/Motivation_crowding_theory

> And we don’t want to use an automated system that selects some ads for us.

So they do direct deals mostly, which is antiquated and probably to preserve their branding as a "tech site".

> Did you know we have a lot of unpaid inventory on Stack Overflow every month?

This is what trying to guess your avails with direct ends up doing. Leaving inventory unmonetized. Typically, you get paid less for less of your inventory trying to do monthly manual trafficking.

All they do is display, so it's rather silly to quibble about CPMs in the 25 cent range...until you look at the traffic and realize that translates to real money.

What you are missing is that there is a cost associated with monetizing the full inventory with automation.

You lose control over what is being shown, you are diluting your brand and you are at risk of alienating your users.

At Stack Overflow we are not willing to make this trade off.

> What you are missing

Automation costs are actually administrative costs (platform fees), which have recently increased via unmonetized opportunity cost models (opportunity being a specific technical term in advertising). SO is not fully or optimally monetizing, probably because they prefer to keep the old tech they are using and/or the staff that has been there since early days. It's also possible they want some dejour/nonstandard controls. Saying they are making an informed tradeoff is simply incorrect. Saying they are making a (unspecified) choice, then not really talking about that choice, is what the article skirts.

Sub 10k/mo for platforms at 2 billion opps per day is AOL and SpotX. That schedule doesn't matter at the volume SO works with. Fees and Opp costs are not the issue with the unmonetized inventory.

Surely you can automate everything but keep manual curation?

Except for all the crap about "targeted advertising", it turns out that if you run a site like StackOverflow, you can more accurately target adverts to your average user by curating the adverts than by letting some system try and guess based on whatever they have managed to sniff.

More importantly, the automated stuff is a huge security risk, and will generally annoy your users when someone puts out something they shouldn't, even if it's not a security issue.

Sure, they could saturate the pages with more crap adverts no one wants, and they would sell for less (because they would be less effective), and they'd piss off their user base.

SO has long been one of the very, very few on my adblocking whitelist. I don't dislike adverts as a core idea, but 99% of advertising online is invasive, annoying and useless.

It's great you are making money. If only there wasn't so many terrible answers ranking no.1 on Google.. :-/

Upon reading the title, I thought they made it possible for users to make money on Stack Overflow.

They did with job placement potential.

Good point.

Thank you for submitting this on HN. I am really happy to see the site grow.

> I am extremely protective of our users. You can ask anyone who works here. It’s is my very strong belief that we have a built up a trust with the community that is not easily earned and is impossible to replace. I want to work for a company that respects that trust, and I do.

> How much money we make is a direct proxy for how much we are helping our users. We focus on maximizing how we help users in order to make a great business. Very few companies have done this successfully, and we take great pride in the fact that we’re one of them.

I can see that they believe in this. As I wrote in my Quora answer https://www.quora.com/Is-Stack-Exchange-still-growing-in-201...:

"They are gaining a lot of goodwill because they have committed to attach a high value to user experience over monetary gain (they learnt from sites in the past that lost their community because they plastered the site with ads. Needless to say, those sites eventually failed)"

I remember other sites that did something similar but they got lost on the way and sold their users away and ultimately fell to disrepair because no one came back.

I hope they continue to stay true to this course.

> We don’t because if we don’t have anything even remotely good to show you, we shouldn’t.

> And we don’t want to use an automated system that selects some ads for us. We looked at this. It didn’t allow us the control we required to maintain the level of quality we want to maintain. We have intentionally left a lot of money on the table. Sacrificing quality is not what we want to be known for. We believe there are better ways.

In my limited experience with online advertising, one metric that almost all advertisers ask for are page views.

I believe this is a fundamentally flawed metric. It can be gamed easily and ultimately targeting ads by page views lead to a suboptimal user experience.

I would rather not show an ad than show an irrelevant ad.

SE is bigger than just page views. It’s a successful ecosystem that has a vibrant community behind it that will grow and nurture itself.

It's great to see that their management would rather not show an ad than show an irrelevant ad.

> If you have time though, I’d appreciate if you told me why in the comments so I/we can do better.

Introduce anonymous answers.

I know of very talented developers who would like to answer questions without setting up (throwaway) accounts. Spam is a non-issue here unless spammers model answers like good answers (of which we have a pretty large set to train from)

Spam has a lot of problem potential. Dodging spam when you're looking for answers is like jumping in a swimming hole where someone dumped a garbage bag of empty milk containers, lettuce leaves, coffee grounds, plastic wrap... Frankly, it's just another kind of advertisement, and I think it's much worse than the official kind, whose site-killing power you mention.

>Spam is a non-issue here unless spammers model answers like good answers

Many people give answers broad latitude as far as "goodness". Users can also be easily confused: http://meta.stackoverflow.com/questions/337843/review-audit-..., and even moderators are sometimes inclined towards leniency: http://meta.stackoverflow.com/questions/333722/how-is-this-q...

A multi-layered system, starting with some filtering right at the beginning, is, I think, the best way to keep illicit ads under control. Stack Overflow has very high value as a target for spammers and very low amounts of spam compared to that value; I'd really like it to stay that way.

I hate having to register at every new site, too, but I don't think it's entirely unreasonable if you're actually trying to post.

For my own curiosity, when you mean anonymous answers do you mean by people who don't have accounts on our platform at all, or do you mean people who already have accounts disconnecting their persona from a specific answer?

You got it in the latter (former would be even better). Explained here: https://news.ycombinator.com/item?id=12961447

> Introduce anonymous answers.

Or, if you are worried about spam, you could allow signed in users to "post a question|answer anonymously". This would not allow users to see who posted the question, but would allow things like statistics (to track spammers and the like).

FWIW, you don't have to create an account to answer questions on Stack Overflow. We'll encourage you to do so, because it generally makes things easier... But if you want to just enter an email address (can be fake) and some sort of name (doesn't have to be one you use anywhere else), you can post without registering or revealing any personal details. We don't require folks to associate an account on another site, or silly "real name" requirements, or anything else that'd force you to identify yourself.

But, I'd guess what you're probably after is something else entirely: a way to have an account, and post answers using that account, but not have those answers associated publicly with that account. This has been requested a few times over the years, but... It's a bit of a troublesome idea:

- Spam / trolling (the kind performed by humans, not bots) often depends on being able to stay just a little bit ahead of the folks who are trying to stop you. Anything that can buy you a few more minutes of activity before being blocked can upset that balance; sure, we could remove anonymity for moderators, but there's a psychological impact here: if you think you can get away with something, you're more likely to try.

- Reputation is tracked on a per-account basis, and most of the accounting there is public: again, this serves multiple purposes, both for the author themselves (they're encouraged to become "invested" in the site by contributing in exchange for privileges) and for the greater community (privileges aren't handed out capriciously, they're earned in ways that can be tracked).

- What about things like edits and comments? Presumably you'd want to be able to respond to comments on your answers, or make edits that correct problems - for "unregistered" answers, that's solved by giving you a temporary account that retains edit rights, comment rights, etc. (at least until you lose the cookie); if you posted anonymously under a normal account, would your edits and comments also be anonymous, or would you show up incongruously as yourself?

- We periodically publish a full set of data on posts, comments, edits, etc. that anyone can use; we'd have to decide whether or not to preserve anonymity in this data, and not just for authorship but also for edits, commentary, etc.

It's a big can of worms, which is why we've so far avoided opening it... But I do see the advantages as well.


- http://meta.stackexchange.com/questions/198/should-there-be-...

- http://meta.stackexchange.com/questions/177154/posting-anony...

Disclaimer: I work for Stack Overflow.

> but not have those answers associated publicly with that account


Joshua, thank you for the response. I was going to reach out and point you to this sub-thread, so you saved me time.

There are some excellent programmers I know who would like to post answers on SO/SE without being able to trace the source of answers to them.

The reasons can be many: some dont want to be bothered with random people tracking them down and contacting them to as involved as having the ability to say (as an example): "ahem, don't depend on this API because it's going to be deprecated in the next 3 months".

I am sure there are other useful reasons.

People might be concerned about misuse and spam, but I have been told Smoke Detector has been doing a pretty good job with that.

A random idea: SO is full of developers who are ready to give back. Why don't you ask the community how they can implement a solution that is not too bad?

Thank you again for the comment. Really appreciate the thoughts.

Ads. Of course. And over half their staff is marketing/sales.


Does somebody has a clue about what % of the revenues come from ads and jobs? I understand that there might be other sources too (documentation and such).

I would pay to get bugs fixed. The snippet editor has been broken for about 6 months. Funny that a site about code would have a buggy code editor.

Would it be thieving, if i made a NN-Crawler, that condensed the Stack overflow-Conversations into a query able chat bot?

Did you know that we provide the complete set of questions and answers (from all sites, including Stack Overflow)?

No need to crawl anything.

You can find it on the internet archive:


And there's also the API - http://api.stackexchange.com

And the Data Explorer (use SQL in the browser to query the Stack Overflow database): http://data.stackoverflow.com

Thanks, thats great..

Seeing their salary calculator makes me depressed what a 1-year experience 1x developer in the NYC metro makes.

Consider how depressed developer salary is in Japan, I'm surprised how well Stackover pays... I'm guessing their Japanese office is tiny.

What do you think is the purpose of this piece?

Seems like it came straight from PR land to me.

tldr: Display Ads and Talent.

> We help companies build their brand and source technical talent through our business products: Display Ads and Talent.

I love Stack Overflow but its community is right now poisonous that i don't want to participate anymore but to be a lurker like many others.

My 2cents.

Re: 'Why we make money'

Is it now shameful to make money, so you have to defend it when you do?

yeah. it's crazy. it costs money to run the servers itself. We are lucky they haven't made the site subscription based or something.

Interestingly, this article was flagged on the new page. I had to vouch for it.

Weird that a previously flagged article made it to the top slot:)

I always thought they earned a nickel for every time they closed a question as "not constructive"

At this point, I have no desire to participate in or help stackoverflow. The attitudes there are such hostile, elitist garbage, where someone is incapable of asking a question, because it will instantly be closed by the mods. God forbid they are a newbie to the site. Responses essentially amount to go RTFM.

This sums it up well.


I used to be very active on SO and I stopped because dealing with new people to the site is a waste of everyones time.

SO has the goal of being a Q to A catalog. You're not contributing unless you are asking a new, answerable question or providing and answer to a new or existing question.

There are tons of people asking questions that are already answered, tons of people asking questions that would take a book to answer, and tons of people asking questions that are subjective and time sensitive. All of this needs to be closed and removed.

The fact of the matter is that tons of questions have already been asked and answered. In fact unless its for a newer technology there are probably not that many left to ask. Because of this about 95% of all new content ought to be removed. In fact you will find a vastly different culture if you stick around the GO tag vs the Java tag.

But the issue is there is a never ending trickle of people who don't realize the above. They sign up and start asking answered or impossible to answer questions, trying to start discussions, or trying to make you debug their specific issue.

You either ignore them or try and help them understand the site by telling them to RTFM.

A more recent problem with Stack Overflow now that its a few years old. The tech gets updated, and previously good answers are now out of date, as libraries and frameworks have changed.

We have an edit button though! And edits are community validated before being merged. I use it regularly to update other people's answers. If the update breaks backwards compatibility, I'll just add a separate section "For Python 3.x, use .items() instead of .iteritems()" etc.

Very true but I think SE does a poor job incentivizing upkeep. Its very little Karma compared to new contributions. Honestly the only thing I with they would change.

I agree, and would like to see a change like that in Q&A, but it would be a pretty big deal at this point in the site's life. However, the new Documentation section does a much better job with this. Each edit is worth +2, and upvotes give points to all significant contributors (+5 or +1 depending on size of contribution), not just the author!

This system has had some issues and may still be adjusted, but I think it's a lot closer to the mark. I've probably gained close to 1000 rep so far from documentation I contributed to but did not author.

> and upvotes give points to all significant contributors (+5 or +1 depending on size of contribution), not just the author

That's interesting. I wasn't aware of the trickle down karma like that.

Keep in mind that's for [Documentation][1], not Q&A (which is what most people think of when they think of StackOverflow).

[1]: https://stackoverflow.com/documentation

Ahh, whoops, you're right about doc karma [1].

[1] http://stackoverflow.com/help/whats-reputation

This is fair. I think it's +2 for an accepted edit vs +10*n upvotes +15 for accepted. Another approach is to just make a new answer on the original question later. I see that a lot for Python 2/3 stuff too.

Your core group of know it all people, won't be around forever. If you keep discouraging new users, by treating them like second class citizens, and a "waste of time". Then the old boys club, is not going to last. I agree entertaining basic programming questions is foolish, but I've seen many examples of legitimate questions, that get the boot because they've been addressed(poorly) previously. At what point do you start just running a Wiki of programming answers, and not a community of programmers. Because the former sounds like what the programming police at stackoverflow want.

I don't agree with your premise. I've seen many users join the site and succeed in recent years, because they cared to understand what they were joining, and how they could contribute to it. If you're only looking for an answer for yourself, people are not going to cater to you. It's not a great environment for random newbies looking for personal help because that was never the primary goal. They're more expected to benefit as consumers of the content that more experienced programmers generate.

From the announcement of Stack Overflow, before the site was even in private beta (https://blog.codinghorror.com/introducing-stackoverflow-com/):

> Stack Overflow is sort of like the anti-experts-exchange meets wikipedia meets programming reddit.

The community elements of the site are nice, and I have made some friends there, but "wiki of programming answers" is much closer to the original vision of the site than "community".

Old boys club is a misleading way to describe the scenario.

There is content that doesn't belong and content that does, this is established by the company at a broad scale and customized for various boards by the community.

I haven't seen the case where content belongs but is denied because you're not well enough know or w/e. If by old boys club you mean long standing members are preventing new members from changing policy then yeah thats true.

Another thing you're failing to realize is that that plenty of people who do RTFM, become contributors, and stick around. More than enough to keep up with demand.

> that get the boot because they've been addressed(poorly) previously

This is a valid complaint. I believe SE needs the Karma system to do a better job at rewarding cleaning up existing questions.

> At what point do you start just running a Wiki of programming answers, and not a community of programmers.

Its more we want a community to create a q/a based wiki. There needs to be a question answering community (no need to focus on the asking one it will come) and strict question moderation helps attract and retain the answering community. Retraining and attracting 15+ year industry vets is much more vital to SE's health than retraining new users.

So your premise is that the site will die because only people who know what they're talking about will use it?

I have to admit some skepticism of this idea.

Exactly. How many times would you want to answer (or even read different questions and answers for) how to compare strings in Java? And it's not limited to one programming language. There was a time that I used to answer but then it started to suck.

That post contains quite a few inaccuracies and misunderstandings about SO functionality.

It has been discussed on HN recently: https://news.ycombinator.com/item?id=12576124

Looks like that discussion talks about the same issue.

joekrie 51 days ago [-]

I find it to be a great resource if someone has already answered your question, especially if I find it via Google. As for asking questions, it's a bit off-putting to be immediately downvoted without explanation, and to not get a single answer.

Point is, it's good resource to free-ride but not actively participate.

city41 51 days ago [-]

That's basically the point of the site. I think SO's owners are ok with this trade off.

FWIW it wasn't always like this. I have a lot of points on SO mostly acquired about 5 years ago. Back then SO was a much friendlier, and much more interesting place to participate in. I haven't actively participated in about 3 years or so.

CaptSpify 51 days ago [-]

I've always thought this described it best: SO is a great place to find answers to your questions, but SO is a terrible place to ask questions.

> Responses essentially amount to go RTFM.

Let's be honest, reading the docs should be everyone's first step before posting a question. So many questions follow the template of "I want to know how to do this but I don't want to put in any effort." The volume of those kinds of questions are what drove me away - not the moderation.

That works with well written docs.

Python standard docs are not great, trying to remember how to use capturing groups on regexes for example (which I use maybe once or twice a year), it takes wading through three or four screens worth of text to get what I want. The Stack Overflow answer is likely to be far more concise, and will give me the example I am needing near the top of the page.

And, on the flip side; people who do RTFM will only ask relatively esoteric questions looking for help on a difficult problem. Good luck getting any answers to that. It happens, but it's pretty rare. Everyone jumps on the latest "why is my hash map implementation crashing?" (It's a bad pointer) in the race for rep.

I am still pretty highly ranked on SO (top 1%), but I haven't been active for nearly two years now.

> Everyone jumps on the latest [bulk question] in the race for rep.

You have just nailed the problem with their gamification.

It takes effort - and thus time - to write a good answer. Since everyone is trying to score the easy wins, you get a flood of mediocre boilerplate answers. In the time it takes you to write a quality answer, the question has been answered multiple times. Likely one of these boilerplates has even been accepted by OP.

For a company that obviously considers the long game over immediate and cheap gains, they haven't managed to impart these values to their userbase.

Incidentally that was also my gripe about SO job ads two years ago. (Iow. when I considered posting something.) At the time their candidate search only allowed to search by answer scores. I had pretty simple needs: What if I'm not looking for developers who can recite trivia? What if I want to seek out the thoughtful ones who actually know how to ask good questions?

Their answer at the time: sorry, we don't have that.

It shouldn't even be that hard: eliminate the insane outliers and jackpots, find the people with matching skill sets who consistently ask decently scored questions.

Bring an alternative to reading the docs is the very essence of SO.

Why read the docs when you can google for the exact single question you have and be looking at the answer a second later?

Especially on technologies you are diving into just to quickly do something, and that you don't have time to learn in any depth.

As someone who regularly advises new users to read the minimum complete verifiable example (MCVE) doc [1], both with review flags and in comments, I really think you should!

Unlike most manuals, it's a fairly short and incredibly well written document. Anyone who reads and follows most of it will be a very valuable user to the community. In fact, following it's process might lead you to figuring out your own questions before you submit sometimes, at least that's how it's worked for me. Sure, it requires effort, but it makes for a better experience all around.

[Many questions that I see get downvoted are authors that make no attempt to follow the guidelines, or have asked a question easily answerable in the standards docs for that language or library. Still, I often answer these personally to provide a bit of hand-holding because what's obvious to someone who's been working with a language for 5 years might not be to someone who's been with it for 1. It's hard to draw a line around obviousness. In any case, if you follow the guidelines you're likely to receive help.]

[1]: http://stackoverflow.com/help/mcve

Are there any Stack-Overflow style Q&A site with communities that feel more like Hacker News(high levels of politeness)?


Wow, what a hostile post. What would you think people would think of HN reading your stuff?

This isn't my stuff. Though myself, and others share the same sentiment. Read the comments below the article.

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact