Faster scanning: We’ve heard that it can feel cluttered or hard to spot relevant content. We believe this update makes it faster to parse a list of questions and identify ones that are relevant to you.

Moving the post score, type, number of answers (if applicable), and acceptance status to the right makes it harder to skim, not easier.

When these stats are on the left, you can reasonably see the score and the title without moving your gaze too much horizontally, but you can just move vertically. With the stats on the right, you can ma-a-aybe see the last few words of the title while you are looking at the score.

I do not see in the examples how answer posts will be shown in lists. But same applies to those - having the type of the post to the left at the start of the title is vastly more useful.

I feel like this undervalues curation.

An old user will know what to do. A new user might not, and worse, not get into the habit of flagging, editing, or closing.

While "less used" - having Edit be an obvious option has a fair amount of value. We want people editing posts and the option to be obvious.

Likewise, flagging is essential for curation, and having it 2 clicks away makes it less likely for folks to.

As much as slimming down the UI is 'nice' - people are not going to use tools they don't see.

I'd also wonder how moving these things will affect keyboard shortcuts for these tools. If you're changing navigation on a page, these might change as well. I currently heavily use them as part of my (tool assisted) moderation and metamoderation workflows.

I find the post summary significantly harder to parse. It's the exact sort of UI that would make me write a userscript to modify it. The biggest problems I see are:

• Distinguishing titles by color is easy to scan. The brain is good at doing that. With the new design, I have to scan based on "slightly larger and more bold", which takes a lot more mental effort.
• The lines between the entries have been removed, which were another subtle visual clue that your eyes would use to locate individual entries on the page. Now, when I read a title and am not interested in that entry, it takes a lot more active reading in order to locate the start of the next list item.
• The top line of each entry is the user name/image, view count, and age. That's the least valuable information in each entry. When I'm looking at this view, I'm most interested in understanding what each entry is about. Whether or not I answer a question is based on whether it's a topic I feel like I know something about. Who asked the question never enters into the equation. At least it shouldn't. When you put the user name/avatar first, it's just extraneous information that I now have to actively filter out. It's now the only colorful part of the entry, so the eye is naturally drawn there.

The new Question page suffers many of the same problems.

• The lack of bounding lines means it's easy to read past the end of an answer and into the comments without realizing it.
• Placing colorful user information at the top of a drab, black-and-white answer draws the eye to the least meaningful part of the answer.
• Placing the "Accepted Answer" flag next to the user info like that makes it look like it's talking about the user and not the answer.
• Someone with a username of "Accepted Answer" and a profile picture of a green checkbox could mislead lots of users who aren't familiar with the details of the new UI.
• There's not much room (especially on mobile) to add a second user's entry above the answer when they've edited it.

In general, I think the elevation of the user's identity to such a place of prominence is a significant mistake. On this site, it's the quality of the content that matters. Who wrote that content is largely irrelevant. I don't care if it's your first day here or if you're Jon Skeet. If you provide a good quality answer, you get an upvote. That's the way it's supposed to work. When you elevate identity like that, you'll force people to start paying attention to it and thus factor it into things (whether they realize they're doing it or not). People may get discriminated against because of their username or avatar. Users with low rep may be less likely to get votes because people see the low rep and then skip the answer without giving it a chance. Lead with the content first, then put the attribution after the end. Literature has done it that way for centuries, and that's on purpose. Let the reader evaluate the content on its own merits. When you put the author first, you're priming the reader's mind with expectations and presuppositions. You do a disservice to both the author and the reader.

Have you done much user-testing of wireframes yet with first time users? Even just some quick and dirty "guerrilla testing" with 2-3 random people like your partner / relative / friendly shop or canteen worker, etc?

Because while there's a lot of nice stuff aesthetically (e.g. I like the neatly contained voting pills), looking at this has me worried:

First of all, I have no idea if the "Reply" button under the question would create a comment or an answer (I guess it's a comment, because I don't see another way to comment?).

But more importantly, I'd bet that most new users seeing this would understand that the "Reply" button is the way to reply to a question (why would there be some other, better form of reply hidden further down the page?), and so of course, these "replies" under the Reply button must be the only type of content answering the question; so:

• if they know the answer to the question, they'll use the "Reply" button to post it instead of looking for the answers box (i.e. answers in comments, which is bad because it further breaks the "beat answer floats to the top" USP)
• they'll get bogged down reading question comments (whereas we typically skim over the clearly-secondary comments to the clearly-primary answers, unless a particular comment stands out). While important, question comments are usually not very useful to people in problem-solving mode who landed here from search (e.g. "is this on topic" debates, clarifications of minor details of the asker's scenario, jokes, opinions, asides, etc)
• many will never scroll down far enough to see that there even are another type of "replies" called answers. They'll see the replies end, then there's a break, another big vote chip that looks like the question at the top of the page, and it looks like a change in topic

The distinction between small, disposable, easy-to-ignore comments and big, detailed, quality-controlled answers was a huge factor in Stack Overflow replacing forums as the place to solve coding problems, and is a major part of the appeal of the wider network for those of us who use it. Being able to skim past pages of the "Why are you even using that, bro" nonsense that plagued forums and get straight to an on-topic, well-structured couple of paragraphs that actually solves the problem was an absolute revelation. I'm worried this will be a big step in the wrong direction and in future, the quality answers (if there are any) will be buried in a nested comment thread, smothered by the kind of fluff we thought we got rid of.

Also, while I've never formally UX-tested SO or SE, I have watched first-time-user colleagues try to figure out the less-famous technical SE sites, and you'd be surprised how many people don't even make it to the answers section in the current designs. I've seen people click onto an SE page that I know has the answer to their problem, start reading the question, and give up around paragraph 3 or after skimming the question comments because they wrongly assumed they'd landed on a personal blog or Medium-like site, where one person was venting about their problem into the void. This makes it look even more like that.

Why would a first-time user keep scrolling down, once they've read all the "replies"?

And there's a bigger, more fundamental question here - why is it that in recent months, the company keep blurring the line between comments and answers, when their separation is a big part of the network's USP, and at a time when your new big competitor (LLM chatbots) are popular because they serve straight-up answer-like content literally modelled on SO answers, with all the comments removed?

Maybe I should ask that as a question.

...because it's a follow-up question, and as we all know, follow-up questions are best asked as new, linked questions so they aren't lost halfway down some other thread and get eyeballs from more people with interests more directly relevant to the follow-up question (we do all know that here, don't we, staff?).

Edit

It is already quite difficult for new users to find the Edit button because it is not a button, but just a word. This typically encourages the following:

• posting code samples and clarifications in comments, not infrequently split across multiple comments due to length limits;
• posting samples and clarifications as answers;
• asking a new question to replace the old one, with or without deletion of the original.

Making the Edit button less visible on the grounds that it is less used is ... a little bit strange. If it is even less visible, I'm sure it will be even less used.

As somebody who has found the edit button, I use it a lot. Posts of new users frequently require markup changes to format code correctly, for instance, or to make pictures show or to correct links.

From time to time, moreover, mass editing has been required on TeX SE and small groups of users have been responsible for correcting hundreds or thousands of posts after a previous site update rendered working code samples unusable.

Making it harder to improve existing posts is a really odd thing to do. Improving both questions and answers should be regarded as a high-value activity. It deserves to be accorded a prominence which reflects that value.

Remember that the majority of these improvements (at least on sites like TeX SE) are made by people who do not get points or badges or privileges as a result. Making it more difficult for them to help the moderators sends an unfortunate message.

Flag

Similar remarks apply to flagging. I do not use this very often, but there are times when I use it repeatedly because, say, somebody is continuously creating new sock puppets or continuously posting spam.

Moreover, flagging should be easily visible, regardless of how frequently it is required. If somebody becomes abusive, it is important that a current visual element clearly signals a way to request moderator intervention. That should never be hidden in some further menu, regardless of how frequently it is needed, because it is of great importance it be easily accessible when it is needed. A new user, especially, who cannot see this option at once is more likely to respond, potentially escalating the situation, or to simply leave.

Hiding the share and edit feature (though in your images share is shown) feels like a very poor idea; I would say that edit and share are items I use very often. Editing a post to improve it, especially soon after it's posted, is really important to ensure it's well received; hiding that feature away is going to make it harder for users to get an author to improve their post when they can't even work out where the feature is now hidden away.

On the new filter UI, I hope is supports wildcards; staging ground doesn't and it's frustrating to use for someone who has a lot of wildcard in their custom filter preferences.

I have to say, I despise "Reply" as a prompt to comment. I hope this isn't a done-deal that I missed along the way. I guess it's an outgrowth of the stated desire to favor discussion (implicitly: over SNR)?

I appreciate that you've laid out the many, varied changes in a thoughtful way. I appreciate that you're bringing ideas to the table early. Make chat work better, make workshopping areas... just please don't turn a high-quality reference library into just another yammering, threaded, internet pandaemonium.

Elevating content creators: We've moved the author's information to the top of the post to better highlight the value our content creators bring to the community.

I disagree. You've removed everything except the name. The current way it's done shows user reputation as well as badge counts. The new one hides a small avatar and user name at the top of a post.

Additionally, with the ad, most users will ignore information around the ad. The avatar is above the title of the question - which the user likely saw either on the page that brought them here, or in a search result, so it's ignored. It's also close enough to the ad that I ignored it on a first and second pass when looking at the screenshots.

If you want to elevate the content creators, don't shrink their name to be the size of the tags and remove their reputation and achievements.

Could we have the big search bar back?

We have a lot of knowledge already on the site, we must encourage users to find it (instead of making duplicates). Design can help with that: A big central search bar emphasises the importance of that frequent use item.

It should not be tucked away to a corner as with the current design proposal.

Below is a rough draft of how I think it should look like:

SO specifically seems to be in need to shift to "curate and update" instead of primary making new content. Therefore not only the search but also the edit & flag function should be put more in focus, not just in optimizing the workflow but also in the design philosophy (see also Larnus' answer, Journeyman Geeks' or cfrs' answer).

While I get the attempt to make comments a first class citizen - many of the current initiatives, removing visual barriers and making them the same size as regular posts does not really work well in practice for a regular user.

I do think these are smaller, but consider how long comment threads are clutter, could we have a way to 'fold' them down if you're removing visual separators from them?

I don't know how other people feel but I think we should try to keep user reputation and badge counts. If not on the question list page, then at least on the individual question pages. For one thing, I think having badges and reputation acts as a indicator of experience. The second reason is that the question page is one of the only places where reputation and badges are visible. If we are not showing them anymore, then what's the purpose of them?

I think there is enough room to show an abbreviated rep count (i.e. 1k) and badge counts inline between the display name and the relative post time.

There are some words there. Most of the highlighted ones, I can barely see - two are basically invisible. And the highlighting blends in way too much with the text.

The code highlighting in dark mode doesn’t have nearly enough contrast, and is way too dark.

Why are the voting buttons so... fat? The background doesn't need to be that big.

Making the custom filters not a vertical list any more, but a series of badges, is horrible as an overview.

The custom filters in the mockup are is not even sorted alphabetically. And even if they were, a vertical list is easier to skim for updates.

Adding a new custom filter will also not cause the "Jenga tower" of custom filters to reshuffle and then push things on new places. If a filter was on the third row at the end, it might now end up on the start of the fourth row.

This decision comes from our comments experiments where we saw a 30% increase in first time commenters.

Currently comments

• are subject to deletion at any time for any reason or for no reason at all.
• can't be edited after a certain time.
• can't be edited by other regular users (unlike questions or answers).
• are often cleared and put into chat for being off-topic.
• are often deleted as unkind.

Note that for many of us, a 30% increase in first-time commenters sounds like a bad thing. Comments are like bad questions, something of which we want less, not more. We don't want you to encourage comments. We'd rather you discouraged them. They're second class citizens for exactly that reason.

On the one hand, we have community standards, which say that most comments should not exist. On the other hand, you're telling us that you are deliberately trying to increase them.

I seriously believe that you need to think about the kind of site you want. Then build that and call it anything but Stack Overflow and Stack Exchange. See how it goes.

As it is, you seem to be saying that you hate everything about the current site except its past popularity. You seem to be hoping that it will recover its popularity if only you completely change it while keeping the existing community. However, you hate everything that the community wants.

If you have a brilliant idea for a new site, why not create it and gather a new community that matches it. Sell the existing site to someone who wants it to raise the funds.

I don't think that remaking the site as an inferior version of Reddit is going to work. Your community doesn't want that, so we'll naturally keep pushing the opposite direction. It would be much easier to create a new site that does what you want. Then you'd only need to convince people to use it.

Improved visibility: Applied filters are now more clearly displayed so you can easily see how the list of questions is being narrowed down.

I honestly cannot see how it is clearer. From the mockup, there are two dropdown menus instead of the single "Filter" menu. Currently, it can be hard to see if you have a filter applied. For example, when the filters are applied to unrelated searches that is unintuitive and unclear because the filter menu is hidden.

However, I cannot see how the new design improves on that. As long as the filter menu (either new or old) is collapsed, you cannot see what restrictions are on the question list. And the design makes it seem like the filter menu will not be open all the time. It looks like a dropdown menu that you open and dismiss as soon as you are finished with it.

If anything there should be an indicator that a filter is active - something that is not easy to miss. The mockup does not show anything like this, as far as I can see.

We’ve heard that it can feel cluttered or hard to spot relevant content. We believe this update makes it faster to parse a list of questions and identify ones that are relevant to you.

As a long-time user, allow me to rank the elements of the question summary, from most important to least important (roughly):

• Score
• Whether there is an accepted answer
• Title
• Number of answers
• Tags
• Excerpt
• Author's reputation
• Modification time (especially unimportant on the recent question list, where we already know the question was modified recently, and on most sites is ordered by modification time)
• Author's name
• Number of views
• Author's avatar

I'm sure other users will disagree about the relative order of a few of these, and it's hard to give a single ranking of importance when question summaries can appear in different contexts. (Remember when the "new questions" and "search" pages had different views which were optimised for what you were looking at? That was nice. Alas.)

But what I must object to is that in the new design, the four least important elements are now the first things you see, and the two most important elements are now over on the right where they are easy to miss.

Post Summary

The first thing I notice is the blocky design. The UI is also a lot larger, and takes up more room. I've also noticed a few special "tags": advice, and tooling. Stack Overflow is supposed to be a library of coding information, and not a forum. Asking for advice or which tool to use is out of scope for Stack Overflow, and should be asked on Software Recommendations. Without the line breaks, the "faster scanning" does not work, as you have to process a dividing line between each question in your mind. If we are going to be in control of density, then please show us some of the less dense options.

Updated Filters

I don't know if this was just cropped out of the screenshot, but the Newest, Active, etc buttons seem to be missing. The new UI doesn't look to bad, and looks less confusing because of how you separated the filtering and sorting, but it also could be seen as more confusing due to the two filter buttons, and the stacked rectangles icon next to them. You could probably have one group of 3 buttons, the filter options button, the sort options button, and the go button.

The Questions List Page

There's a lot to go over here. I'm going left-to-right and top-to-bottom.

Top Bar

There is no use to have the AI Assistant linked in 2 places, and this is just a little confusing. The Create button looks quite forumish, and should be changed back to Ask Question.

Left Column

First of all, it's nice to see that you can collapse the side bar without going into settings. This is a lot easier to find and it could feel more familiar for new users if they can easily collapse it to see the main content better. I think the AI Assist logo is quite similar to the Super User logo.

The "Communities" section. As I said before, this is not a community site, it is a library of coding information. This looks very much like Reddit, but that's not necessarily a bad thing. The UI is nice, but you don't want users thinking that this is a forum, not a formal Q & A.

The content section looks nice, all you changed was making it collapsible. The resources section is also nice, but I would move it above the "communities" section, as you will are more likely to be looking for one of these things than a different "community", but it is nice where it is.

Middle Section

I already went over most of this, but this shows better how it is stitched together.

Showing how many questions have been asked is really not necessary, and anyone who needs that can get it through the API. It does seem that there are only three buttons for the filtering, but this only brings more confusion to how to use it. This is also the first time we've seen an ad here, and I think it is very dirty and deceptive to have the Reddit-style ad hidden in there like a post. If there are ads, they should be obvious, and not in the way of anything.

Right Column

I think this looks nice, and this is a better spot to place an ad.

Redesigned Question Page

Having the question info at the top, by the title is somewhat nice, though not having an edit button, and not showing who edited it last and when could be slightly annoying. (btw, how come in the old version all the editors have This_is_not_a_forum's PFP??) On dark mode, the code block's highlighting scheme most definitely does not meet regulations for contrast.

Having the questions designed more minimally appears more like a forum, and less like a Q & A.

On the right side bar, the "Python Newbies" community looks quite like a collective, and I do not think we need 2 collective systems. If you are renaming all the sites to communities, then having 2 community definitions could definitely be very confusing,

The community activity sections looks out of place, and more like a marketing tool than something useful. Excluding chat from the side could be a mistake, as then people will only ever find the lobby, even if they are over 20 reputation.

With the related questions, you seem to be working backwards, since you now include the poster's name and profile picture.

See this for more about why Stack Overflow/Exchange should move away from forum-style features.

In the question list, some question are marked [tooling] or [advice]. Those look like special (meta-?)tags (though [tooling] already exists as a regular tag). Where do those come from? Are they automatically added by e.g. an AI module in the system, and if so, what are the criteria for this?

I'm a big fan of the new longer post summaries, but can I convince you to make the light-theme body text a fraction darker? It feels pretty low-contrast and borderline hard to read, which is maybe ok for the truncated summaries, but really not great for the actual question page.

(I realise that font rendering is fiddly, and I'm very willing to believe that part of the problem could be e.g. a rescaled screenshot.)

1. I hope (and mostly assume) that the New Post Summary is two alternatives pasted together, and that either you'll choose one or we can flip between them. That "repeating but in a different format" (without a clear break to say "this is an alternative") was pretty jarring.

2. I usually get annoyed at sites that shorten summaries too much, so that I have to click through to see even the full headline. But for this site, title only is probably sufficient if you don't trim that title to fit in whatever size visual box you were hoping for.

3. Most sites doing visual design (as opposed to just semantic markup and letting the browser defaults apply) worry way too much about getting the box sizes pixel-perfect, and not enough about readability. If I have to increase the font size (and I often do), I would much rather have a box that is "too big" than have the content clipped and unreadable, or (slightly less bad) unreadable without a horizontal scroll and scroll back for every single line.

   This is even more important on phone screens, where scrolling or zooming only a portion of the page (usually either fonts in general, or code examples specifically) is often worse than with a keyboard and touchpad.

4. This might be a quirk of your screenshots, but it looks like you actually shrunk the fonts, rather than making them larger. (Maybe it is just unbolding?) I prefer the New Posts summary with just the titles (though it probably depends on community and even individual poster), but the first alternative was clearly worse than the current format, at least in that picture, when viewed on my screen.

5. The "you said you made fonts bigger, but they look smaller" problem also happened on the filters and sorting, but there it might turn out to be OK, since that is mostly text that won't change very often. If it is a popout or separate window of some sort, though, no need to keep it small.

6. I can't actually read the (screenshots of) the Question List or New Question pages. But that is also true of the "current" format, and I can read the current site, so that is probably a screenshot issue.

7. It would probably be helpful to show the same content in the new and old formats. Though if you do that, be extra wary of getting to the point where you only look at the visuals and don't bother trying to read them. A visual map with tiny or blurred words is sufficient for navigation when you are familiar with the page (as you will become for any particular test page), but not sufficient for users trying to read a particular post for the first time.

8. A wider default is probably good, but please ensure the page still works on a phone or in a tall, narrow window off to the side of whatever someone might have been doing as their main task on a larger screen.

To follow up on peterh's objection, I propose to remove the empty spaces on both sides. It is not obvious on the OP screenshot, but a pretty large empty margin is present also on the right side.

You can also return the scores to the left. The page is easier to scan with this items on the left, like a dotted list.

Here is how it looks like.

Are the sort options currently saved as part of a "saved" or "custom" filter? How is that preserved/changing/etc. now that the sorting is done in a separate dropdown menu?

Disclaimer: I personally do not use that feature here, but if I were a committed answerer or bounty hunter, I imagine I'd want to keep the sort with the special filtering.