User activation: Learnings and opportunities

Question

We mentioned in our roadmap update that user activation was an area that we are prioritizing. In this post, we will go through the background, challenges, and opportunities that we will be focusing on.

Background

We believe that Stack Overflow and the Stack Exchange network are an essential platform for learners and experts to come together to share and create the highest-quality knowledge. Over the years, developers have asked millions of questions, answering which has only been made possible through input and insights from the community. We believe that participation matters because the more users find, ask, answer, and curate content, the more everyone in the community learns and shares knowledge.

User activation means introducing newer users to the platform in order to help them accomplish their goals. It represents those early touchpoints, from the moment a user signs up to when they start using the product in a way that provides value to them. You can think about the steps to product usage in these phases:

1. Acquisition - the ability to get users to sign up
2. Activation - the ability to help users realize the value of a product, which converts a user from a passive visitor to an active participant
3. Engagement - effectively engaging with the product

Activation is the gap between sign-up and engagement. It’s where we (and most platforms) see the biggest drop-off. We see this as the biggest opportunity for improving user retention. Focusing on initiatives related to activation means helping users get the most from their experience. Our prior efforts around sign-ups allowed us to get those users in the door, and now we want to help activate those users to become valuable members of the various Stack Exchange communities here.

What is the challenge?

We believe that more participation is key to success. However, one challenge that we currently face is that users are trying to participate, but can’t. We have a few hypotheses as to why that is:

• Most users don’t currently have any or enough reputation
• It has become increasingly more difficult for users to gain reputation, and without reputation, there aren’t many ways to interact with the platform beyond asking or answering a question.

Hypothesis: Most users don’t have reputation

When we look at how reputation is distributed among our registered users, we can see that it’s heavily skewed towards low rep, which was also true before the recent signup initiatives.

Hypothesis: It’s become increasingly more difficult for users to gain rep

If we look at how much reputation has been rewarded to all registered users over the last few years, we see that there’s a declining trend. There are a few possible explanations for this, one being that it’s hard to ask a high-quality question that hasn’t already been asked. Another is that it’s hard to find relevant questions to answer. We have ongoing efforts to address the question-asking experience through Staging Ground.

Hypothesis: Users are trying to participate but can’t

We then examined how often users tried to vote or comment. This percentage represents attempts to take these actions from users who do not yet have the necessary privileges.

What is our opportunity?

We want to help new users recognize the value of the product and build the foundations to lead these users to eventually become part of the community: we want to find ways to get users familiar with the site, begin to teach these users about community norms and site mechanics, encourage them to engage with the site by beginning to take actions, and offer ways for them to be a part of the active community.

We’ve previously shared some work around sign-ups and experiments around tags, as well as a request for feedback on what a homepage should look like. We hope this gives additional context as to what types of initiatives we aim to prioritize in the future.

Next steps & feedback

The upcoming projects we are working on:

• Experimenting with tags
  Hypothesis: watching tags is a lower-effort action that will allow users to more easily customize and find relevant content.

• Homepage (starting with newer users)
  Hypothesis: Providing a starting point to learn about the platform and a home base will provide value to users by educating them about different areas of the site and delivering more personalized content.

A potential upcoming problem statement we want to explore further is that users don’t take actions despite having the required privileges because these actions are placed in different areas of the page and because the page doesn’t differentiate between which they can and can’t take. You can expect more communications around this when we start diving deeper into this problem area.

If there are any other problem statements you think we should look into, or if you have any thoughts or feedback on the ways we can encourage newer users or users without reputation to participate, let us know as an answer on this post. We will be monitoring this post until September 25th.

Answer 1

Problem number one

You changed signup which resulted in huge amount of accounts created and now you are complaining that those users which were not interested in becoming users in the first place are not active enough. Related Dramatic increase in New User accounts - why?

Problem number two

Again you got the order of usage/participation wrong.

1. Acquisition - the ability to get users to sign up
2. Activation - the ability to help users realize the value of a product, which converts a user from a passive visitor to an active participant
3. Engagement - effectively engaging with the product

The above order is nonsense. People are using and can recognize the value of the sites without having accounts. This is how the sites are meant to be used for vast majority of people. The primary use case for repository of knowledge is searching and reading, not active participation.

So trying to "activate" all those users which never wanted to have an account in the first place will be mission impossible.

You need to show value by keeping the reputation of being the place where people can easily find valuable answers. If people keep coming to the sites because they can find answers for their problems or other interesting information, they might join some day and start actively participating. But you cannot force that. It has to come naturally, and not everyone will want to participate.

Problem number three

Your emphasis on the acquiring new users at any cost is making you additional problems with attracting users that may turn to be active participants. Namely, you have turned the sites into a spammy looking and annoying popup sites with the signup pop up.

Do you know what I am generally doing when I land on such site which shows me unprompted signup popup? I close the page and never return again.

Nobody was having trouble with signing in. It is imaginary problem people don't have. You just had plenty of users that never had the need to sign in.

And even if you get such unwilling users to sign up before they are actually ready, whatever they do, they will not be able to start because they require reputation for everything, except asking or answering.

And if they start asking or answering unprepared, they get immediately punished as it is very likely that not being familiar with the site and the rules they will break some.

The actual problem

You do have problem with participation and engagement. And some of it comes from the inability of new users who want (or have been lured in) to participate to do that successfully. Lack of reputation is not the main issue here, although access to some features like chat and ability to post comments that are tied with reputation would help for some of them.

Main issue, which also reflects in lack of reputation, is that new users are not educated well about how sites work and what is expected of them. They are not clearly presented with rules and consequences (post ban) of not following them. Many also don't know that there are multiple sites in the network and they ask questions on the wrong site.

Better onboarding, showing more detailed instructions than the tour has and less verbose than what is in the Help center would also help. If nothing else, clear and to the point explanation of what kind of questions are on topic and which are not for particular site would go a long way.

Most new users are also not familiar with editing system and the fact they can earn some of the reputation by editing and improving posts. I learned that good edit can earn you 2 rep on SO, only when I crossed 2000 rep mark. It is silly, but it just shows how even users that don't have too much trouble starting will need time to get to know the system.

Another problem you are not considering here

You are focusing on attracting new users (which is something that is needed), but at the same time you as company are doing everything in your power to drive existing long time users away. Many of those that have left the sites or reduced their participation are experts and curators that are providing the most valuable product: knowledgeable answers and curation of bad ones. Without those, you have nothing.

Suggestions

• stop driving existing users away and start listening to the moderators and the community
• remove sign in popups as they serve no real purpose
• make it easier to find good answers on the sites by improving search
• revisit Outdated answers project, especially answer versioning part
• give better tour to new users and initial guidance so they are not surprised by the rules and at the same time that will make it easier for them to earn initial reputation
• this also includes better guidance about using AI when posting as many don't know AI posts and even partial AI posts are not allowed (banner on the answer editor is too little, too late, and can be easily dismissed)

There is probably more...

Answer 2

I'd like to share my perspective as a newly "activated" user:

First as a student and then as a very junior developer and sysadmin, I have always enjoyed SE (SO, superuser and serverfault, mostly) as a source of solutions. I did not feel confident I was competent enough to provide my answers to anyone else's problems, and critically, I did not have any problems that weren't already solved by SE.

There's 2 things in what I just said. the first is that I believe SO, superuser and serverfault are mostly solved, in that either my issue is so specific that the answer won't be very useful to anyone but me, or the issue is so widespread that my question is already answered. This isn't to say that the site is useless or that new questions are useless either, it's just that as a new user, you basically have no way of contributing meaningfully to these sites. The fact that when you have an issue and try to upvote the answer that helped you, the site tells you to go away (because you have 1 rep), and so you do.

The second thing is that as a non-activated user, the only actions I may take are post questions and answers, or leave. This isn't conducive to me feeling like I'm part of any community at all (and in fact I believe that without chat.SE, I still wouldn't), and just makes me want to come for questions already answered, take my code snippet and leave. I don't think removing these restrictions makes sense either for moderation reasons, but I'm not pretending I have all the answers.

What I'm getting at is that without a paradigm shift of the entire platform, the want to "activate" users out of sheer love for reading stackoverflow questions is nonsensical. Users that "activate" do so out of their own will, because they have found something to come back to, and I genuinely don't think "finding SO questions to answer to" as a day job is something to come back to. Personally, it took me finding codegolf.SE, which I grew attached to because I am autistic, then getting chat privileges introduced me to an actual community of users, and I come back to see what they have posted. SO, SU and SF in my opinion are not conducive to finding a community like puzzling, codegolf, worldbuilding are.

In my opinion, if you really want more registered users more active more of the time (why?), the big technical Q&A sites are not how you convert them, and smaller more community-oriented sites are a significantly better tool to do that, through showing the diversity you offer and letting a would-be convert find their niche. I am sadly acutely aware that the endgame of what I'm describing is ultimately just old reddit without the cat videos, maybe I just miss it.

Thanks for reading my ramblings, good luck in the meta.SE jungle.

Answer 3

I absolutely agree that users are trying to participate in many ways but can't, which is why I've championed the idea of moving away from the reputation system as a method to award privileges. Over the years I've learned that users want to edit more freely, review posts, chat, etc...

One reason I'm glad to know y'all are working on improving search is that it's another place where users have indicated the are struggling to participate. Users have often talked about how they rely on Google to find anything on SO because finding an answer to their specific question in the millions of posts is so frustrating. As a result, they often ask questions that have already been answered because it's easier for them to ask a new question rather than find an existing one.

While Staging Ground is a good start and I hope to see continued work there and on the question asking process, I'd also love to see a better answer writing flow for first-time answerers and people who have had struggled writing good answers previously. I can't overstate the value of "just in time" user guidance - where we build in help as a user is trying to use a tool, rather than trying to pre-inform them about something they're not trying to do (and may never do).

Most frequently, we all know that users want to vote and comment - likely in much higher volumes than any other privilege. This is understandable - the mental barrier to vote or comment is likely significantly lower than the one to ask, answer, or edit for many people. Besides in your vote and comment attempt data, we can see this in the huge number of answers moderators and reviewers on SO delete every day that are attempts to comment or show appreciation.

Unfortunately, the risks within the current system of allowing everyone with an account to vote and comment are high and addressing them is a complex Gordian knot that can't simply be cut through because it happens to be holding up the Sword of Damocles, so it's reasonable to look for other places to focus.

User personalization and privileges

As to the next steps you've listed - I understand personalization is a big thing right now across the internet and can be an effective way to increase sign-ups. Just look at the boom the days after Dark Mode was introduced on SO. Control over homepage content is one of the consistent ways users have asked for more personalization, so I understand why that's something you're investigating and, since tags would be the most obvious ways for users to restrict what shows up on their personalized homepage... I can understand why that's part of it.

That said, while SO's volume is so huge that users customizing the home page isn't likely to impact moderation and may improve it, I'd encourage you to investigate the risks for smaller sites where the homepage is a de facto review queue. For example, if many of the active users customize their homepage and don't check the full feed, visibility of low quality content could drop and leave more closing to fewer users - particularly moderators.

This may mean it's even more important to consider visibility of moderation tools for users who have privileges - whether they use them or not. There are absolutely under-utilized tools that could use a visibility boost but I do request that you do so in the interest of discoverability and ease of access, rather than pressing users to do things they're not interested in doing. In my experience, users have a relatively limited list of activities they're willing to do and most people with any given privilege have no interest in that aspect of the site.

The review queue indicator is a case study in frustration and sorrow that ended in a fizzle that still needs to be "fixed". It's annoying to people who have no interest in reviewing while simultaneously under- and over-notifying users who are interested in reviewing, leading to many reviews aging away or taking a long time to be attended to. Happy to recommend some posts to review if you are interested.

Deep dive into voting data on SO

Assuming I'm reading your third chart correctly, you're indicating that if you combine all votes and all vote attempts, the latter makes up about half of the total, potentially doubling the number of votes cast. Assuming that many people don't even bother to try voting because they know they can't, those numbers are probably lower than they could be, indicating a huge opportunity for engagement.

Looking at the data on Feedback attempts VS Votes

Using a couple of public SEDE queries Jon Ericson wrote for an answer last year, I looked at the vote and vote attempt history for Stack Overflow. One of the queries looks at votes by type per day on any given site; the other looks at feedback by type per day on a site.

Because I couldn't get the queries on a single graph, I pulled the CSV for each and moved it into Google Sheets. I also shortened the window to start in 2021 to get a view of more recent history. As a note, there are elements in this result that lead to questions about whether it's measuring what I think it does, so if you're able to share your info or confirm this is correct (or at least the same data y'all are using), that'd be helpful.

Here's what that looks like (Stack Overflow between January 2021 and today) - the faint lines are centered, moving average trend lines for each:

As the chart above shows, while registered vote attempts are practically non-existent (assuming the data is correct), unregistered vote attempts have been about equal to actual votes as of June 2023, though before that, they were quite a bit lower. This was a bit confusing to me since the question is about "activation" (getting signed-up users to participate) but the numbers I'm seeing in SEDE that align with your chart represent users who are not logged in. Hopefully you can share data that shows the correct numbers for registered, low rep users.

Since the registered user vote attempt count in SEDE is so small compared to the others, I'll include the a chart of it on its own.

There are a couple of questions that these charts elicit that the one in the question doesn't -

• Why has there been such a drastic reduction in votes from users who have the privilege, despite the stability of unregistered votes?
  • Does this change represent a smaller number of users casting votes? fewer votes/day per user? both?
• Why was there such a huge jump in downvote attempts from unregistered users in June 2023?
  • This appears to be the related to the vote button styling changes. The test for this change was held in June 2022, coinciding with the jump there, too, including the change having an outsized impact on downvotes.^§
• Why are there so few feedback attempts from registered users? Is there a bug in the way the data is collected and/or de-duplicated or do they just not try to vote because they know they can't - or something else?

There's one last thing I thought I'd draw attention to - the percentage of votes by user type, which I did by dividing the upvotes by total votes for each user type.

Questions this chart leads me to asking include

• Why do unregistered users attempt to downvote at a rate that's nearly split 50-50 between up and down while registered users primarily upvote?
  • Is the "cost" to downvote answers part of it?
  • Does being anonymous lead users to try to vote down more frequently?
• Why did both unprivileged groups see a reduced percentage of upvotes in mid 2023 (and was it really related to the vote button update)?

I'd love to have some insight into these questions and many others. In your post, you say:

Activation is the gap between sign-up and engagement. It’s where we (and most platforms) see the biggest drop-off. We see this as the biggest opportunity for improving user retention.

Considering the platform has lost more votes per week over the last few years than there are vote attempts per week from unregistered users, I don't really think you're correct. I think the biggest opportunity for user retention is retaining (and getting back) the once-engaged users you've already lost by understanding why they left or reduced their engagement on the site.

Now, sure, there are millions of never-"activated" users out there, why should you chase after the significantly smaller number of people who left? - because the odds are probably still better. 100% of the people who left were previously "activated" to some degree. What percentage of visitors have even attempted to vote or comment, let alone tried to participate regularly? I don't know if that's even quantifiable but the general community industry wisdom I've read is that it's close to 5-10%.

What counts as "activation"?

When I see definitions of "activated" and "engaged" referring to community products - particularly Stack Exchange - that imply "activation" is key to users finding value in the product and that activated/engaged users can't be "passive visitors", I'm concerned ^‡. A book that isn't checked out of the library may still be read by people in the library. If you're trying to justify investing in the library, you can't only consider easily-countable things like check-outs. The goal has to be finding ways to identify passive use and showing it has value - though, if you can solve that, community managers everywhere would love to know your findings.

While giving passive users ways to participate that make finding information or customizing their views easier is great, I wouldn't expect it to go any further than that - and I don't know that y'all have any data that shows it does. For example, of users who created accounts when Dark Mode was released, is their engagement with the platform significantly different than users who created accounts at other times?

That said, active users who create, curate and vote on content are absolutely necessary to the success of the platform, and there aren't nearly enough of them. I'd argue that if the goal is to have a sustainable platform full of high-quality content, you should be looking to "activate" users who show they have interest in tasks that need more engagement and use just-in-time guidance to lead them through it without getting in their way more than necessary to get them up to speed.

I don't think that attempts to vote (or even comment, for that matter) on their own should be considered indications of an "activated" or "engaged" user or that a user is ripe to become "activated". I also believe many engaged users never vote or comment. As such, only citing voting and comment attempts as support for the general hypothesis that "users are trying to participate but can't" leaves out a lot of engagement while also overstating the potential benefit to the site. ^†

Conclusion

I appreciate that y'all are focusing on features for registered users. Make the site easier to use and customize in ways that people are telling you they want; remove barriers to participation and ensure easy access to all features, particularly those that are under-utilized or forgotten. Feel free to look for better ways to solve existing problems or even remove complexity that isn't doing what it was designed to achieve or is no longer necessary but ensure you fully understand why things work they way they do as the obvious reason for complexity is rarely the only reason.

Additionally, understand why people are reducing or stopping their engagement. Are there things that can be changed to draw them back or prevent others from leaving? Someone who leaves because they're tired of waiting for a ill-functioning system to be improved may return once it's fixed. Newly "activated" users won't be interested in all aspects of the platform, so may not necessarily replace the users who left. Prioritize "activation" efforts to areas of the site that need the most help.

If reputation is truly the barrier, look towards trying to understand alternatives. In an era when rep was easy to get, it was an easily-surmountable barrier for most users to reduce problematic behavior on the site. Rep is no longer easy to get, so it has become an insurmountable barrier for many users. While you can't just dump it, and there likely won't be an "easy" solution, we don't have to accept the status quo as the only way but the risk of problematic behavior is still there, if not greater.

Find good solutions to the privileges problem, not the easy ones.

---

Footnote:

§ - There is an increase in upvotes but it is difficult to identify whether that's due to a short-term downturn in anonymous voting in April and May of 2023, as the weekly vote counts in late 2023 are consistent with historical voting from this group. The increase in unregistered up and down votes during the test in 2022 appears to encompass all of the "net 28% increase in overall votes" reported when the change went live in 2023, as there's no similar jump in actual votes. That said, the first half of 2023 showed an increase in voting decline by privileged users. While there wasn't an increase in votes for those users, the decline slowed somewhat, though it's not possible to draw any conclusions about why.

‡ - "Why 95% Of Your Community Visitors Don’t Participate (And What You Should Do About It!)" is an article from Feverbee - a community consultancy and training company - that makes some really great points about "lurkers" - who they prefer to call "learners", which I think applies to this situation really well. In it, they share that most lurkers don't participate more often because they don't feel like they have anything to contribute and that onboarding messages don't work while efforts to pressure them into participating more frequently usually end up backfiring.

† - For reference, see the issues that plagued Interpersonal Skills SE for the first year of launch due to "drive-by" votes, comments, and answers from users with the association bonus who had no interest in sticking around to participate otherwise.

Answer 4

I don't really see or understand how experimenting with tags or the new user homepage helps with any of the hypotheses. If Stack thinks the barrier to new user activation is rep... how would changing the home page, or making tags popups... "different", affect any of those?

I do agree with the hypotheses #1 and #3 (they are directly linked), however the only attempt at solving this problem I've seen in recent past was shut down due to a negative reaction within the community. That said, I don't think just making rep or privileges more available will solve the larger issues the site has. It might prove to increase user activation, just as being annoying improved user signups, but because the clash between being a knowledge base and the help desk new users tend to treat the sites as would still exist and user will still feel like they're walking on eggshells when just looking for help.

I mostly disagree that it's somehow harder to earn rep today. Yes, users who were active earlier have far more rep typically, but I've yet to see any evidence it's due to the questions being asked today vs a decade ago. If anything there's more opportunity today to earn rep than in the past given there's less answerers generally. I don't buy into this whole... everything's already been asked and answered idea. The overwhelming majority of questions asked on SO are debugging (help desk) questions that don't neatly match with any given duplicate and are generally easy to solve... the same kind of questions that make up most of the 2k+ answers I have.

---

Generally I feel like the biggest hindrance to people joining and participating with SO is the atmosphere, the general vibe of the place after a decade of neglect, of dealing with the same problems year after year with no end in sight. Constant hate from the outside because people tried to use the site in a way it isn't designed for. If most new users who come here are coming here for help and the site isn't designed for help of that nature... I don't see how moving privileges around or changing rep levels (or making cosmetic changes as suggested) would do anything about it. People will still ask "low quality" questions and get shut down rather than helped and the cycle continues.

Answer 5

I felt something odd about it for a while, but couldn't put my finger on it.

Then I read this comment by curiousdannii:

"Activation" is a really dehumanising

And it hit me. That was nagging me all along. And the puzzle parts clicked into place: the core problem all along, since the management got eh, refresh, is that this new management treat the users as a product.

And that is a problem that can't really be solved.

Changing the name e.g. "User participation" would be now only sugar coating.

Answer 6

We then examined how often users tried to vote or comment. This percentage represents attempts to take these actions from users who do not yet have the necessary privileges.

Out of curiosity, how was this calculated?

• What is the percentage out of? Total number of users?
• Does it do anything to normalize for users that haven't been active for a long time?
• What about users with no account?
• What if a user tried the action multiple times?
• What if the same network user has multiple network site accounts? Is each one counted as a separate account? What if they have different privilege levels on different sites?

Frame challenges

My experience and observations of posts and other user accounts aligns with the hypothesis that it's harder to earn rep (compared to ~10 years ago), but... I also earned ~40k rep (essentially from scratch- for the record) on SO in the past ~2 years, answering both old and new questions. I recognize that earning rep isn't the same across tags (more people use VS Code than CMake, so I get more upvotes on VS Code answers than CMake answers), and the qualities of Q&A aren't the same across tags... but it was at least possible.

I'd like to challenge how much it matters that it's harder to earn rep, and whether there are other ways to look at the problem.

• The basic actions of asking, answering, and suggesting edits don't require rep. That's a lot of opportunity to contribute to the value of the platform already.

• From my observation both on-site and looking at off-site commentary on the platform, a lot of people who want to comment but can't are

  • looking for updates on a Q&A (they should follow the posts, and maybe share the Q&A in a related external space), or
  • seeking detail/clarity on a question (there's a good change the question should be closed), or
  • seeking more information about an answer (technically they can post a new question post, though I have my own thoughts on how ideal that is).

  Adding to that, there is a basically endless stream of comments posted every day where the commenter shouldn't have commented. I have numbers on this, and they're big even with conservative heuristics-based queries.

  What I mean to say is, many of the people who want to comment have no idea what the purpose of comments on this platform are, so handing out the privilege to more people- especially towards a demographic with less familiarity with the platform- is not what I anticipate (given the data and my experience on the ground) to be a particularly net-productive idea.

• Voting is important, but I think it's good for it to be a privilege to stem off vote abuse. I'd rather have

  • Add an option to sort answers by the viewer's previously cast votes (you have the power to make voting more valuable to registered users regardless of whether they have full voting privileges or not),
  • maybe https://meta.stackexchange.com/a/393604/997587 (there are unexplored design avenues to make anonymous votes have relevance to the public audience),
  • and something related to but maybe not exactly like Once I reach 15 reputation, will my old votes be taken into account? (at least being able to see past "post feedback" votes and "re-cast" them as real votes).

  In other words, there are several high-impact (in my opinion) ways to make anonymous votes more useful to individuals and the community both now and later without lowering the privilege's rep threshold or changing what actions result in what amounts of rep changes (which can be disruptive and have unforseen/unpredictable negative consequences).

Lastly, if someone really wants to vote and comment everywhere, even if they don't have the background or the patience to get rep from asking or answering, getting upvote privs is 7 helpful suggested edits, and comment everywhere is 25. I'd guestimate you could maybe reach 50 rep that way in a month or two (depending on context). Also, as an aside, accept vote rep (+15) is a lot at that scale of things.

Answer 7

I am not sure this will have any effect, but I may as well record my nagging here just for fun. Some of you may even notice that some of these are things I have been lamenting for YEARS. Maybe this is the lucky time? Who knows.

As usual, any instance of "you" refers to the company and not to the OP who just draw the shortest straw in the "who is going to post this" contest.

First issue. Incomplete premises

---

Let's start from your "hypothesis"

Hypothesis: It’s become increasingly more difficult for users to gain rep

I am somehow concerned. Either this is an EXTREME case of tunnel vision, or a deliberate attempt at retrofitting the data to the conclusion some executive already came to overnight. In both cases, this just shows that you already have a plan and that you are actively blinding yourself to only see the behaviors and the data that fits your picture.

It is true that gaining rep points is getting harder, but you are failing to see the reasons, again thanks to a good degree of tunnel vision and only focusing on what you already decided the results should be.

There are a few possible explanations for this, one being that it’s hard to ask a high-quality question that hasn’t already been asked. Another is that it’s hard to find relevant questions to answer

You just focus on a saturation of content resulting in an increased difficulty in writing good content. Have you ever considered that people are voting less because the most active curators of the site have been getting increasingly burnt out and fed up with the recent antics and have either left or greatly reduced their activity and this may be a part of the issue too? And before you answer "Yes, our Stack Overflow centric data shows that" ... have you ever tried to get some data that is NOT Stack Overflow?

Let me tell you something. The reputation system is bad. It always was. It is not even a "Prosus meddling" issue. It is simply broken. The numbers do not work for anything that doesn't reach SO core mass.

I don't think most people seem to get this. Not even employees. I have tried multiple times to raise this issue, but I have always got a deaf ear. I don't know why this is so hard to admit, but at times it almost feels like there is a lingering ... arrogance almost? ... that everything that works for SO is perfect and must be the same for every site.

I started my journey on the network on SharePoint. Back then I used to read the question list daily, always looking for something I could answer. All my efforts gave me nothing. I don't even have the 10k required for moderation tools.

I posted this picture on the Tavern back in 2021. And it wasn't even the first time I rambled about that.

This is a screen from the SharePoint badge page. I don't know if you realize...

SINCE ITS CREATION BACK IN 2008, LESS THAN 1K QUESTIONS MANAGED TO GET MORE THAN 10 VOTES, AND ONLY 3 GOT ABOVE 100

Do you want to compare this with SO?

Want to talk about the top users instead?

Currently the top user on SharePoint is Waqas Sarwar MVP♦ with a "huge" 56k rep ... remind me ... how many users over A MILLION does SO have? NINE, with three more getting very close. I am currently sitting in the top 60 of the whole site. I have 6.8k rep. My last answer dates to FOUR YEARS AGO.

Do you think that maybe something is very wrong here? Do you think that maybe users can feel insulted and unrewarded when they curate for years and can't even get the "trusted user" privilege? SharePoint has only 12 users you "trust". And some aren't even active anymore.

I remember that SO always celebrated its users. Correct me if I am wrong ... there was even a special celebration for reaching ... 100k? Something about a special shirt or a plaque. I have a SharePoint shirt. But I got it through other means. For full disclosure, Rebecca once held a custom swag giveaway on SharePoint that was re-calibrated to the site numbers ... but for some odd reasons I wasn't included. I never managed to get an answer why.

Anyway, the point here should be clear. You have a system that heavily favors SO and basically tells everyone else to go [redacted] themselves. I wonder ... do you think Waqas Sarwar with their 58k on SharePoint has any less dignity or showed any less effort than the top users on SO? Did you ever tell them that? Did you ever think that maybe more users on your low traffic site deserve the fuzzy feeling of being called "trusted"?

The takeaway of all this should be likewise simple

All users are equal but SO users are more equal than the others
Animal Farm The Stack Exchange Network

So, to summarize, the actual reasons it is getting harder to gain rep:

• content saturation
• less curation, the more you burn out your users the less they will vote
• gaining rep has always been harder on some sites, yet not only the network doesn't account for that but the gimmicks in place also actively work to rub said issues in your face, constantly reminding small site users that they are "second class citizens" whose efforts are given far less dignity that many "do my homework" users get on SO.

Second issue. "Activating" new users, alienating the ones already here

---

It is not a secret that the company is struggling to get new users on board. Yet for all the focus on getting new users to join, to the point of literally showing them begging pop-ups at every click or immediately recreating accounts they just deleted (something that reminds me of those ADS filled sketchy "free" mobile apps) the same effort doesn't seem to be put into RETAINING your existing users and more importantly ... making sure they enjoy their time here.

Once again, I don't get the reasoning behind all of this. Doesn't the company really not realize that the shrinking userbase and contribution levels aren't just justified by "less new users resources" but also by a profound wound that is bleeding out the contributors they already got? Or do they think they can "replace" the old dissenting voices with new and fresh ones more willing to follow their marketing goals?

Does the company think that sitting out on serious issues their users pointed out will work to increase retention (I for one am still waiting on the answers about the serious privacy concerns you promised about 2 months ago)? Does the company even realize just how mad their users are? I propose a challenge for you. Publish a new survey with the following questions.

• On an 1 to 10 scale with 1 being "totally bad" and 10 being "awesome", how do you feel about the Stack Exchange company?
• On an 1 to 10 scale with 1 being "totally bad" and 10 being "awesome", how do you feel about Prosus?
• On a 1 to 10 scale, how willing you are to suggest Stack Exchange to a friend or coworker?
• On a 1 to 10 scale, how likely do you think you will be a user on the network in 6 months from now?

... hopefully you get the idea. Go on, I challenge you. Just remember to make the results public, we don't want to miss out ... and look on the bright side: after all, the odds are in your favor, anyone still willing to answer a survey probably still has far more goodwill for you than average. So, whatever bleak result you get should still be far better than the actual reality ... I mean, I doubt you will get the users that already moved to Codidact to return here and answer a survey just to let you know how much they dislike what this place has become.

I must be honest, I told you this enough times already and I really can't understand if you think you know better than us or you simply don't care. Alas, my suggestion is that if your are bleeding out it is probably wiser to first stop the wound before starting a transfusion, unless of course you want the new blood to soon join all the rest that you have already spilled on the floor. Not only will it help the numbers in the long run, but you will also put a stop to all those tears of blood that got fed up and left just to become evangelists of the "Stack Exchange is a horrible place owned by a horrible company" propaganda. I told you before, go look at the comments you get on reddit, twitter, discord and so on... then return here and think again about your plans.

Third issue: users are a resource to spend to get content

---

Knowledge has always been the focus of the network, the goal its users aspired to. Yet, over the years the values pursued by the users and the company have come to diverge. On one side the users, still aiming at building a library of knowledge that everyone can use. On the other the new owners, desperately seeking some profit and therefore seeing the knowledge as a precious sell-able product that they want an exclusive distribution license for.

Coming from this, it is easy to see why the company is so focused on finding a way to legally own said knowledge, constantly devising new restrictions to gatekeep its distribution.

At the same time, it is similarly easy to see why users are seen as a resource before being actual people. Just on this post you got three? users telling you that talking about "users activation" is somehow insulting and it is actively dehumanizing your userbase. Yet this too you don't seem to realize. I am not saying this is made on purpose. Yet I can't help but think that such slip-ups still simply betray what the company view is anyway.

You just want to increase your raw materials so that the sites factories may produce more finished product content that can be sold as LLM training material. But the users are not just ores that you can put in a furnace to smelt and get some ingots.

They are users, they are people. So, if you want them to join then you really have just one option. Have them enjoy their time spent here.

Answer 8

Your reputation graph does not necessarily show that it’s harder to get reputation. Earning reputation takes some amount of effort, so it could show that people just don't care enough about reputation to take actions that would improve their ability to earn it, like responding to feedback to improve their posts (editing also improves the chances of ending up as an HNQ), keeping one account instead of just creating an account when they need one, etc. That could be because they don’t care about the privileges reputation gets them, or because reputation just isn’t perceived to be worth as much as it used to be. Having 3000 reputation on SO now doesn’t make someone stand out as much as it did 10 years ago when fewer people had tens of thousands of rep.

You have to make people want to be part of the site before they will care about badges and reputation that only have value in the context of the site. This is a community building problem, not a gamification problem. Frankly, I think dropping the reputation requirement for chatting or commenting would have a much bigger impact than no rep voting. There is no way for people to ease into the community to learn the norms. You have to post something well-received to get reputation to be able to participate in discussions with other people or you have to lurk without participating for a long time. You’re setting new users up for failure by making their first interactions with the community have to be an answer or a question.

I have under 200 rep on Stack Overflow, over 10k on ELL and over 20k here. I used SO, but I didn’t contribute there. Not because it was hard to get reputation, but because it was hard to get connected into the community. Gamification will work up to a point, but you still have to give people a reason to play the game, and to keep playing the game after they’ve earned all the points.

And as an aside, new users don’t have to ask a high quality question that hasn’t been asked before; they have to ask an interesting question that is reasonable quality and hope that the right people see it before it scrolls out of view. People answer and upvote duplicates all the time. There’s no reason that the goal should be one decade-old question that has The Answer.

Answer 9

The same feedback always gets brought up over and over with voting in that there always seems to be a lack of insight into teaching new users what should be upvoted (or downvoted) according to the guidelines shown in the tooltips. Without this, theres a risk that an apparent successful campaign to get more engagement will not accurately reflect quality.

So I applaud the attempt to find new engagement but please make sure that it isn't just a hollow metric.

Answer 10

I am convinced that both for Stack Overflow and for the common good of the internet, there is nothing more important than keeping the active "experts and curators" happy. New users are not useful unless they ask questions that the experts consider to be worth answering or are an expert themselves.

Yet, as a company, you seem to care more about maximizing the number of "users", "page views" and hence short term advertising income than you care about keeping the "experts and curators" happy.

Until your aims as a company becomes the same as the aims of the "experts and curators", the company will continue to be seen as the problem and not a provider of solutions/value.

Answer 11

There's a lot of opportunity for improving user engagement that isn't covered under the aegis of "activation".

Why don't I use the review queues on Stack Overflow (SO)?

I'm glad someone asked (yes, I realize that it was me). The primary reason that I don't use the review queues is that I have no clue whether the questions/answers are good. I mean, if the answer is "Hey, check out my fake viagra at this link" then I could be pretty sure that it was a bad answer. But for answers that look responsive, I don't know if they aren't because there are a lot of subjects on SO where I am not even an amateur much less an expert.

If the review queues showed me only posts where I had some level of expertise, I would participate in reviews. That same discovery mechanism could also be used on the front page to show me only questions that I am somewhat likely to answer. The review queues provide better feedback though, as the standard is to review all the posts. If I'm skipping them, then that is more useful than the few questions that I'll choose to answer from the front page.

You may wonder how I find questions to answer now. I mostly don't. The majority of the questions that I've answered have been ones where I wanted the answer and couldn't find it. In the end, I wrote the answer that I wanted to find. As such, I primarily rely on Google to discover questions, and I mostly answer questions where I am not a subject matter expert.

Answer 12

I think you guys are are taking the right steps trying to increase Acquisition/Activation/Engagement for new users. I'm not critical that way unlike some other users.

But, in my experience, it took me 10 years to sign up out of privacy concerns. (In the meanwhile I dictated a few Q&As that my colleagues posted for me and one is at +600 votes. All so I wouldn't create an account and get caught up in the site.) When I finally did sign-up, it wasn't long before I realized my privacy concerns had been well founded all along - some people called me stupid for saying we needed GDPR nearly 20 years before we got it. So here's the catch: you can get more users to sign-up and do stuff, but without a clear privacy policy defending users these measures are just band aids that won't solve the long-term underlying problem.

users don’t take actions despite having the required privileges because

And you can count me into that group. Not because I don't know where functionalities are, but the company has been taking the approach of solving community problems through business process improvement. In some cases it will work, but some core issues will need a sociological formulation - I haven't seen so much as an attempt at that.

Answer 13

it’s hard to ask a high-quality question that hasn’t already been asked

That's great news! It's evidence that SO is succeeding at its mission.

(I'm primarily focused on StackOverflow. I'm open to the possibility that I might have different opinions in the context of some other Stack Exchange sites.)

Having a lot of accounts being created that then go dormant does sound like a problem, but I highly doubt the solution is that SE needs to put effort into "activating" these users.

Consider the three primary use cases for Stack Overflow:

1. Somebody has a question, finds an answer on StackOverflow, and goes on being productive.

   There's no reason to sign these people up, let alone activate them or worry about how engaged they become. In fact, it would be a detriment to all parties.

2. Somebody has a question, doesn't find an answer on StackOverflow, so they create an account to post a question.

   A subset of these people aren't on the right site, and they don't figure it out until they post their first question. It doesn't serve anyone to try to activate these users.

   Another subset of these people are students taking their first and only programming or computer science class. They aren't developers and they aren't planning to become developers. Chances are, their questions aren't even new; StackOverflow and the search engines simply failed to connect them to an existing question with good peer-reviewed answers. Trying to activate them is a waste of effort. Spend that effort on trying to improve the tools that help users find the existing questions.

   The rest of the folks in this use case will probably eventually find themselves in the next use case.

3. A developer wants to answer questions, vote, and/or help curate.

   Maybe it's pure altruism, maybe it's a way to pay-it-forward for the help they've received, maybe they want to show off or gain visibility with recruiters.

   If these people are creating accounts and not activating, that's worthy of attention. But my guess is these are the users who are activating.

Before you decide the goal is to increase user activation, consider whether the underlying problem is that too many unnecessary accounts are being created.

Answer 14

Nitpick since you're asking for feedback: It would be good to not round to 1% in your histogram but leave an extra significant digit. 1% could be anything from 0.5% to 1.49% which is a rather broad range for the 5 out of 8 bins. I know that the emphasis of the plot is on the 87% of low rep users, but nonetheless would be good to have enough precision for the rarer reps.

Reproduction of image for ease of reference:

Answer 15

There's 2 concepts that I think need focus (and the lack of, or any counter-measures will have a negative effect)

1. User ability to accomplish goal(s)

   Fundamentally a user is using software to accomplish a goal (or set of goals)

   For StackOverflow/StackExchange it is either to solve a problem, earn 'Fake Internet Points', abuse the system to promote/spam some other endeavor, or a combination of these

   • If trying to solve a problem, any barriers to being able to post the question, post code related screenshots, or add comments... will cause dismay.
   • If trying to earn reputation (which includes trying to prove your good intentions), any barriers to helpful participation are super frustrating.

2. User sense of belonging/community

   Belonging is a fundamental human need, deeply ingrained in our psychology. Maslow's hierarchy of needs places belonging and love right after physiological and safety needs.

   As a new member to a community, you want to feel like you belong... and the more you feel like you belong, the more you in turn will want to contribute back to the community. If your home page improvements include very clear "Here's the things you can do to start" callouts then this may be quite helpful.

   • Tied to #1 above, if I struggle to accomplish my goals or get hand-slapped in the process I'm not going to feel welcome
   • (I'm a long time user, so I don't know this, but...) Does StackOverflow/StackExchange provide onboarding options that let a user participate in a healthy way that lets them feel welcome and appreciated?
   • Negative feedback... I understand the reasoning behind not allowing upvotes/downvotes to users with little to no reputation, but this feels overly heavy handed. I can add items to my cart on any shopping site, and deal with the signup part later... as I have a 'vested interest' in participating. Could we still provide the messaging that the votes don't count (yet)... but acknowledge that the user has voted for the question/answer (with all the visual updates) advising them that they will be applied when reputation levels have been reached. e.g. We welcome your contributions and will keep them safe until you can fully participate.

Finally, I noted that there are evil doers that want to spam the system, or abuse it to promote whatever other thing they have. We do still need to block this as it negatively affects everyone on the site, but I think we have to be careful with how we treat new users in the "gray zone" after sign up, where we don't know if they are a good or bad user. I'd prefer where possible to presume they are innocent until they prove otherwise.