As an answer to another question I wrote out an analysis that, effectively, Stack Overflow is likely close to the point of saturation. Consider the following:

According to Evans Data Corporation, there were 23 million software developers in 2018, this number is expected to reach 26,4 million by the end of 2019 and 27,7 million by 2023. - DAXX

If the 26.4m estimate is correct then Stack Overflow has roughly 42% of the world's professional software developers with 11m users. To put this in context, Facebook has 2.45 billion active users which is about 32% of the world's population (their target user base). While Stack Overflow says their target user base to be "anyone that writes code" I'm not really convinced that you are going to get most of them as registered users. Case and point, in my professional community I know a number of people who write code and will occasionally look things up on Stack Overflow, but don't see the point in asking questions since the answers are already their, or the active user base is on a listserv.

As a result, it appears that if the company wants to see overall market growth they need to focus less on programming (i.e., Stack Overflow and affiliated network sites) and more focus on building the features and user base to support other sites like Arqade (148k users), Personal Finance & Money (64k users), or Worldbuilding (58k users) to name a few. When you consider how many people play video games, it's shocking that there are only 148k users on the site when a Q&A based video game site seems like a great idea.

As such, how does Stack Overflow plan on increasing traffic and the user base of other Network sites?

  • 13
    Seeing as users past a certain reputation see less ads, lurkers are more profitable than registered users. Nov 30, 2019 at 21:14
  • 2
    There will be new questions with new technology. One small example: the rise of Docker creates lots of questions which wouldn't exist a few years ago.
    – dustytrash
    Dec 1, 2019 at 1:31
  • 1
    The Hot Network Queue serves as cross-site advertising. We can talk about it, if it is useful or not.
    – peterh
    Dec 1, 2019 at 2:14
  • 1
    @Renan Although, more high quality answers to more questions => more google hits => more lurkers
    – Rakete1111
    Dec 1, 2019 at 11:01
  • 1
    SO ties job postings to the Q&A website. The same could be done with other SE's such as Tech support & serverfault
    – dustytrash
    Dec 1, 2019 at 17:35
  • @dustytrash what about Christianity or Mythology? Dec 2, 2019 at 2:13
  • 3
    You mean the network consists of more than SO? </sarcasm> Dec 2, 2019 at 11:41
  • 4
    Although I can agree with the core of your request, the sentence "If the 26.4m estimate is correct then Stack Overflow has roughly 42% of the world's professional software developers due to the 11m users." is completely false. You realize that those 11m users aren't all professionals, but could as well be hobbyists or students, for a very large part? Or duplicate accounts because somebody forgot his crendentials? Or dead people, or sockpuppets, or spammers, ... Pretty sure that if you make a survey across all developers, you will get much much less than 42% of them having an account on SO.
    – dim
    Dec 2, 2019 at 12:46

3 Answers 3


In summary, their strategy is to just keep things running as long as there are ad-clickers around or the user base of possible potential future ad-clickers is growing. Regardless of if the site is SO or a small beta site that was just launched.

The ads

  1. Having a high bar for question quality drives newbies away.

  2. Newbies are also driven away from SE's due to some veteran users' behaviour and actions (like closing and downvoting bad questions, will revisit this soon from #16 on).

  3. Having more newbies around means more clicks on ads.

  4. More clicks on ads means more money.

  5. So, solving #1 and #2 is a way to have more #3 and reach #4.

  6. Another way to have more clicks on ads (and thus more money) is to simply push in more ads.

How the values of the company changed

In the beginning:

  1. SE's founders valued question quality, signal-to-noise ratio and having experts around.

  2. Transparency, democracy, freedom of speech and a stake in decision making and governance is something that experts value.

  3. Having a lot of experts around is the way to build up valuable questions and answers and have people clicking up on ads.

  4. Question quality, signal-to-noise ratio and a core of field experts is the differential that allowed SE outperform its competitors at the time. Losing them would not allow that.

However, as SE matured:

  1. SE grew to the point that there weren't any important competitors anymore, establishing itself as a monopoly in the internet world.

  2. The question quality fell slowly but progressively over time since some point around 2014, dragging with it the signal-to-noise ratio. But regardless of that, most experts stayed around anyway (at least before Monicagate) and no significant competition grew.

  3. Since there is no meaningful competition anymore, the experts have nowhere else to go in order to have a meaningful participation, so #8 is not needed anymore.

  4. Since there is no meaningful competition anymore, the newbies have nowhere else to go in order to get their questions satisfactorily answered, so #7 is not needed anymore. But they still can leave the site nonetheless.

  5. Newbies just want their question to be answered ASAP, so they don't value what experts value in #8.

SE behaviour changed

Let's remember:

  1. So, solving #1 and #2 is a way to have more #3 and reach #4.

And since we have #13, let's see about #1 and #2 in closer detail:

  1. Newbies are driven away by having their questions closed, downvoted, deleted and/or receiving harsh comments.

  2. Review queues UI's is optimized for people reviewing questions and answers quickly without leaving useful feedback.

  3. Newbies don't understand what MCVE actually means, that name is awkward.

  4. Newbies are also driven away by SE's unwelcoming toxic environment (the "no soup for you problem").

  5. People participate more if they earn reputation. But newbies earn reputation more slowly.

  6. Some groups of people don't participate because they are excluded somehow. If they don't participate, they won't click ads nor generate revenue.

So, how to make newbies feel more welcome?

  1. To solve issue #18, how about renaming it to min-reprex?

  2. To solve issue #19, let's create a Code of Conduct to ensure that people know how to not be "unwelcoming" and moderators know how to enforce it.

  3. To solve issue #19, let's add a bunch of "user12345 is a new contributor. Be nice, and check out our Code of Conduct." messages everywhere.

  4. To solve issue #20, let's raise the reputation gained by question askers (a large part of them are ad-clicking newbies).

And what can be done about #21?

  1. SO don't seems to have many women around. Let's investigate why!

  2. SE does some surveys on that. However they commit some flaws in their surveys that lead to biases and misinterpretation of the data.

  3. They concluded that women feel excluded by the community and that it should be because those sexist bigots excluded women.

  4. They failed to see that although women really did feel excluded in SO and that there are indeed a few sexist jerks out there posting horrible comments, the reason why women feel excluded is (mostly) not because veterans are bigots actively excluding them, it was simply because the relative proportion of women interested in programming AND also in participating in SO is for reasons that severely exceeds SE's small world smaller than the proportion of men, and so more women than men are excluded because of lack of interest in being included. So, they saw a correlation but misidentified the cause.

  5. Since they "discovered" that women are excluded by those bigots at SE, it seems reasonable to seek for other excluded people.

  6. Lgbtphobia is a thing that happens out there and occasionally here in SE (there are always a few jerks around). Also, since there are some LGBTQI+ people on staff and on moderator board, they really feel that pain.

  7. Someone referred to a woman as "he". Either by accident and naiveness or for real malice and jerkiness.

  8. Someone else referred to a queer or non-binary with the wrong pronoun. For roughly the same reasons as #32.

  9. From #32 and #33 misgendering is a cause of #19 and #21. Since there are staff members among those misgendered people, the complaint found an easy way to SE's management.

  10. From #34, the Code of Conduct should be changed.


  1. SE started to talk about lgbtphoby and misgendering and how to change the CoC (#35). Remember that #8 is not needed anymore as per #13.

  2. Aza had some trouble about misgendering with someone somewhere and resigned.

  3. Due to #37, SE's management panicked about #34 and #36.

  4. Monica was in the wrong place at the wrong time and was cast as one of the culprits.

  5. Monicagate happens.

I.E. That all was triggered by a combination of factors that are highly improbable to happen individually and just cause an effect when happening together. However, as statisticians know, sometimes the Murphy's law acts up and in some random unlucky day, the terrible combination of bad factors indeed happens and triggers an explosion.

MSE is going to die

  1. Due to #13, #8 is not needed anymore.

  2. Due to #40, almost nobody in MSE is sympathetic with SE.

  3. Almost no ad-clicking newbies use MSE. Because it is not a useful site for newbies.

  4. Since #8 isn't needed anymore and we have #42, so nobody from management will give any useful feedback whatsoever for anyone on governance questions here after the last CoC updates. In other words, they won't listen nor talk to anyone here.

  5. However, since MSE is very much a toxic community, people might still get offense from it.

  6. From #45 and #34, we see that LGBTQI+ people might take offense if MSE is not closely policed.

    • To be fair, in fact everybody who is participating in MSE at those troubled times will eventually feel offended by someone else. Multiple times.
  7. The most that SE's management do is to order CMs and mods to keep the flag queue under control and suspend whoever crossed lines, but due to #44, they don't want to hear anything more about that.

  8. Monicagate took a legal turn.

  9. So, from #48, MSE is not just toxic, it is actually damaging!

What we have in MSE?

  1. Bug reports are on MSE and are useful, but MSE sucks as being a ticket tool. Let's create a proper ticket tool.

  2. MSE is the way to get user feedback and to post announcements. However, since we have #41-#49, it ceased to be.

  3. Self-governance are also part of meta. An advisory board of moderators might replace that.

  4. Support is also part of the meta. However, SE can manage it without involve the community.

  5. From #53, SE introduces the loop.

  6. From #41-#54 MSE would not be useful anymore.

So, finally we reach our near future:

66. Execute order 66.

What more?

  1. Careers is a way to generate revenue and is sitting down in a dusty corner. But since we have #13, there is no risk.

  2. Since careers face strong competition from elsewhere, it is improbable that investing too much on it would make things significantly better.

  3. The teams thing is a new cool stuff that SE wants to monetize.

  4. Teams don't need the large community anyway.

  5. Let's recruit a new CEO to make teams happen. He doesn't need to have any stake in the larger community because that is not why he was hired and there are other people handling that.

  6. As Joel left the company (following other people who did the same before), no one who cultivated the values of the newborn company (#7-#10) is still around anymore.

  7. SE is still completely unaware and blind about issue #17. But I have no hope anymore of anything being significantly better even if they try to do something to address that.

Now, I see the big picture!

  • 1
    Is there any empirical evidence for #2? Like, have newcomers pointed out specific instances where behaviour of veteran users made them not want to participate further?
    – MechMK1
    Dec 2, 2019 at 15:00
  • 9
    @MechMK1 Almost every post in every meta about newbies complaining of having their question closed or downvoted telling that this should be because there is some crazy conspiracy going on against them or because people hate them for no reason. There are thousands of examples out there. I am not saying that the veterans did something wrong, but that are many times that newbies naively thinks that they did. Dec 2, 2019 at 15:02
  • 5
    Then I would interpret this less as caused by the veteran's behaviour, and more as a direct result of SE, Inc. failing to guide newcomers to write good questions.
    – MechMK1
    Dec 2, 2019 at 17:50
  • My take on this is that they are moving from a 'patriarchy' (or a hierarchy based on competence) to a 'matriarchy' (or a hierarchy based on the needs of the apparently most vulnerable). Possibly more for ideological reasons than commercial reasons (well, their ideology makes them believe that this will be commercially viable). Various influential (or vocal) members of (Western) society have the same goals on a larger scale. It remains to be seen what will become of it.
    – Benjol
    Dec 3, 2019 at 12:37
  • I wonder whether the community partly brought this upon themselves by pushing for more 'social-networky' functionality (comments, chat, etc.). Not that those things are bad in themselves, but they create expectations of 'niceness' which are unrealistic in the context of a "it has to compile" mentality.
    – Benjol
    Dec 3, 2019 at 12:44
  • 2
    @Benjol I wouldn't categporize that as "patriarchal" or "matriarchal" (and the usage of such terms with such conotations are very likely to trigger flagging). Anyway, whatever are the terms used, what I see is some sort of class conflict, before it was "almost everything for the experts and almost nothing for beginners" and now it is "almost everything for the beginners and almost nothing for the experts" - Both models are severely flawed. Dec 3, 2019 at 14:14
  • @VictorStafusa, yeah, the terminology probably is a bit flippant, which is why I used the quotes. As for flawed models, I don't think you can get large quantities of unpaid experts to babysit beginners, in any case not with a system based solely on Q&A. If the time put into Documentation had been invested in a tutorial- or coaching-based offering (or at least an attempt in that direction), I feel that would have been more useful.
    – Benjol
    Dec 4, 2019 at 5:40

My impression - based on what SE Inc. has been doing lately - is that they don't have a plan, that increasing the participation on the other SE sites isn't a high priority.

The thing is, I think the most they could have would be a set of strategies that they think might increase traffic and user base, but with so many diverse communities, they'd need to have different plans for almost every site. The way you attract and retain experienced system admins is almost certain to be different from the way you attract and retain knowledgeable science fiction fans, which will be different from how you attract and retain people who know how to cook...

But before the question of how do you attract the users you want/need, there's the question of who are the users you want/need? Do you need experts? How many do you need? And that ties into what I think is the most interesting question about what SE Inc. is doing these days - are they no longer trying to build a repository of knowledge? If not, what are they going to build?

I see no indication that they're thinking about these foundation issues, so I strongly doubt they have any actual plans that would follow from developing a vision of what they're going to do in the future.

  • 3
    "what are they going to build" - An ad view generator Dec 2, 2019 at 15:06

According to some publicly available data from similar web, only in the last 3 months the SO lost 10 millions in traffic, that is quite a lot considering the short time. IMHO this come as an effect of the unfortunate decision to shut down the jobs section and the story one. I feel this all went to advantage of LinkedIn.

  • 2
    Well - SO careers kinda underperformed over the history of the network, and we'd actually had 2 rounds of retrenchments related to that. Also - SE made google analytics cookies optional, so the traffic that's detected went down. Its not a accurate metric to compare traffic before and after that. Jul 23, 2022 at 13:44
  • Could you include the source of this data?
    – Ryan M
    Jul 26, 2022 at 10:47
  • @Ryan like i wrote simply go to similar web and check the estimated traffic. For my knowledge that is even overestimated. Jul 26, 2022 at 12:29
  • @Journeyman where I compared SO with others network? I wrote about SO with its own traffic. Also those numbers are effected same for any platform, so the analytics topic is irrelevant. Put more clear, if SO loses traffic due as analytics restrictions, that should be same for all others platforms, but this isn't the case as there are platforms gaining traffic. Jul 26, 2022 at 12:29
  • Ah, I didn't realize that Similarweb was a company. The link: similarweb.com/website/stackoverflow.com/#ranking
    – Ryan M
    Jul 26, 2022 at 13:24
  • Wonder why I get rid off -4 points if things have changed. On top I would like an explanation why my comment, reporting simply a trend that got only worst since the first question was made. Don't think I will receive an explanation. Jul 28, 2022 at 9:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .