5 ms·

Given that they always had the occasional, well, quite frequent actually, "all our servers are busy" problem, while having a monetization problem (meaning inves

Given that they always had the occasional, well, quite frequent actually, "all our servers are busy" problem, while having a monetization problem (meaning investing in infrastructure isn't something decided lightly), I find it perfectly reasonable. Or do you have a concrete - and I mean concrete monetization plan for that data? Not just an "idea", it should be as real as the cost of creating and maintaining the additional server(s) for tracking (even more) stuff, and also why it's more important than finally solving their "busy servers" problem. Everything sounds easy from a high-level management perspective. Until you actually have to do it and can't just wave hundreds of "details" aside.

soared 10y ago
That is a good point. I forget how small of a team they have and their existing/significant problems. I guess I should be more amazed at reddit's lack of .. success? They have so many users and so much engagement and yet they tried to monetize with a 3rd part affiliate link program. I suppose data isn't something they could competently use.
- toomuchtodo 10y ago
  Users and engagement don't always make for profit unfortunately (or even revenue for that matter!).
- Klathmon 10y ago
  You also have to take into account the average reddit user. The ones against any kind of monetization at all are already vocal enough, let alone the people who would balk at anything that monetizes user data. If you look at their change where they introduced affiliate link rewriting it's really surprising. They did it in a way that made sure they wouldn't affect any 3rd party apps, they gave everyone 2 ways to opt out, and didn't rewrite links that already had an affiliate code. And yet even with all of that, there was a ton of outcry and anger at the change. Reddit as we know it won't be massively profitable, because the vocal users of reddit are so completely against any form of monetization.
  - gedrap 10y ago
    But people are vocal against pretty much every significant change, aren't they? Think of all the times when Facebook iterated and change the format of the profile. There will always be vocal minority which will never be pleased, no matter what you do. And later they forget it and move to other thing. Sometimes you just have to carry on with your plan knowing that you will lose some users, but aiming to do better overall in the long turn. It's important to listen to the users. What's even more important is knowing when to say 'no'.
    - Klathmon 10y ago
      But reddit's "Vocal Minority" is much larger, and much more vocal than most. Almost to the point that it might actually approach the majority of active contributors. Yeah, no matter what there will be naysayers, but the majority of comments in the announcement thread for the affiliate links thing were negative. Combine that with the reddit demographic's love of a controversy (no matter how trumped up) and you have a ticking time bomb that makes mistakes hurt for much longer and much worse than most platforms. Just look at their previous "algorithm change" fiasco. They changed the front page algo, and after about a week they changed it back because of outcry. For the next several months people continued to complain about how everything was worse now, and that they were lying about "reverting" the algorithm. It's insanity.
      - gedrap 10y ago
        But aren't these horribly entitled and grumpy users... toxic?
        Klathmon 10y ago
        I'd say so, but when they make up a not insignificant amount of your userbase, can you afford to just ignore them?
        derefr 10y ago
        If 60% of your users don't want your company to make any money, maybe you need 60% fewer users.
- collyw 10y ago
  What do you class as success. More users or more money.
ben_jones 10y ago
I'm fairly certain most of Reddit's server problems came from database R/W operations. If you look at some of the metrics they've put out recently you'll see things like "queued upvotes/comments/messages" suggesting that they now heavily queue them. They also cache SUPER aggressively, very noticeable on large threads like those for sporting events where you can refresh a page and not see any updates for several minutes. I think this is a talent problem and not a "throw servers at it" problem. EDIT: As much as I like bashing Reddit I feel like I need to mention something. Reddit is the UNDISPUTED best news source for large, breaking, news stories. Reddit's live threads are curated and crowd sourced feeds containing information from police scanners, News outlets, twitter feeds, and a variety of other sources. It's value is in its ability to display conflicting information instantly such that the viewer can get the clearest picture possible. This thread of today's Dallas shooting is a prime example [1]. [1]: https://www.reddit.com/live/x7xfgo3k9jp7 https://www.reddit.com/live/x7xfgo3k9jp7
- rangibaby 10y ago
  Reddit also gets it wrong. They blamed a completely unrelated person (who had killed himself) for the Boston bombing, which is now a meme ("We did it Reddit!"), and even today a photo of a black man in army camo and holding an M16 was falsely claimed to be a shooter.
  - ben_jones 10y ago
    Absolutely it can be dangerous, such as the use of police scanners (which will often broadcast uncorroborated information).
  - eterm 10y ago
    That photo today was published by the Dallas PD, it wasn't a reddit thing. https://twitter.com/DallasPD/status/751262719584575488/photo/1 https://twitter.com/DallasPD/status/751262719584575488/photo... Besides, you're mixing up reddit the platform/company and reddit the commentariat.
    - fennecfoxen 10y ago

soared 10y ago

That is a good point. I forget how small of a team they have and their existing/significant problems. I guess I should be more amazed at reddit's lack of .. success? They have so many users and so much engagement and yet they tried to monetize with a 3rd part affiliate link program. I suppose data isn't something they could competently use.

toomuchtodo 10y ago
Users and engagement don't always make for profit unfortunately (or even revenue for that matter!).
Klathmon 10y ago
You also have to take into account the average reddit user. The ones against any kind of monetization at all are already vocal enough, let alone the people who would balk at anything that monetizes user data. If you look at their change where they introduced affiliate link rewriting it's really surprising. They did it in a way that made sure they wouldn't affect any 3rd party apps, they gave everyone 2 ways to opt out, and didn't rewrite links that already had an affiliate code. And yet even with all of that, there was a ton of outcry and anger at the change. Reddit as we know it won't be massively profitable, because the vocal users of reddit are so completely against any form of monetization.
- gedrap 10y ago
  But people are vocal against pretty much every significant change, aren't they? Think of all the times when Facebook iterated and change the format of the profile. There will always be vocal minority which will never be pleased, no matter what you do. And later they forget it and move to other thing. Sometimes you just have to carry on with your plan knowing that you will lose some users, but aiming to do better overall in the long turn. It's important to listen to the users. What's even more important is knowing when to say 'no'.
  - Klathmon 10y ago
    But reddit's "Vocal Minority" is much larger, and much more vocal than most. Almost to the point that it might actually approach the majority of active contributors. Yeah, no matter what there will be naysayers, but the majority of comments in the announcement thread for the affiliate links thing were negative. Combine that with the reddit demographic's love of a controversy (no matter how trumped up) and you have a ticking time bomb that makes mistakes hurt for much longer and much worse than most platforms. Just look at their previous "algorithm change" fiasco. They changed the front page algo, and after about a week they changed it back because of outcry. For the next several months people continued to complain about how everything was worse now, and that they were lying about "reverting" the algorithm. It's insanity.
    - gedrap 10y ago
      But aren't these horribly entitled and grumpy users... toxic?
      - Klathmon 10y ago
        I'd say so, but when they make up a not insignificant amount of your userbase, can you afford to just ignore them?
        derefr 10y ago
        If 60% of your users don't want your company to make any money, maybe you need 60% fewer users.
collyw 10y ago
What do you class as success. More users or more money.

ben_jones 10y ago

I'm fairly certain most of Reddit's server problems came from database R/W operations. If you look at some of the metrics they've put out recently you'll see things like "queued upvotes/comments/messages" suggesting that they now heavily queue them. They also cache SUPER aggressively, very noticeable on large threads like those for sporting events where you can refresh a page and not see any updates for several minutes. I think this is a talent problem and not a "throw servers at it" problem. EDIT: As much as I like bashing Reddit I feel like I need to mention something. Reddit is the UNDISPUTED best news source for large, breaking, news stories. Reddit's live threads are curated and crowd sourced feeds containing information from police scanners, News outlets, twitter feeds, and a variety of other sources. It's value is in its ability to display conflicting information instantly such that the viewer can get the clearest picture possible. This thread of today's Dallas shooting is a prime example [1]. [1]: https://www.reddit.com/live/x7xfgo3k9jp7 https://www.reddit.com/live/x7xfgo3k9jp7

rangibaby 10y ago
Reddit also gets it wrong. They blamed a completely unrelated person (who had killed himself) for the Boston bombing, which is now a meme ("We did it Reddit!"), and even today a photo of a black man in army camo and holding an M16 was falsely claimed to be a shooter.
- ben_jones 10y ago
  Absolutely it can be dangerous, such as the use of police scanners (which will often broadcast uncorroborated information).
- eterm 10y ago
  That photo today was published by the Dallas PD, it wasn't a reddit thing. https://twitter.com/DallasPD/status/751262719584575488/photo/1 https://twitter.com/DallasPD/status/751262719584575488/photo... Besides, you're mixing up reddit the platform/company and reddit the commentariat.
  - fennecfoxen 10y ago