5 ms·

This is spot-on. HN has generally massively over-estimated the amount of porn on Tumblr, in terms of percentage of total content. I say this as a primary source

This is spot-on. HN has generally massively over-estimated the amount of porn on Tumblr, in terms of percentage of total content. I say this as a primary source with first-hand factual knowledge.

coldtea 7y ago
It doesn't have to be a large "percentage of total content". It just has to be the content that matters to most users...
- evanelias 7y ago
  Well, "most users" implies a majority, which is simply not the case here, full stop. Far from it, especially if you define a user as a person who has a Tumblr account and posts content on the site. If you consider a "user" to include non-monetizeable lurkers who were just there to view adult content, then saying "most users" is definitely still wrong, but perhaps slightly less so. Tumblr's peak was many years ago (2013 iirc), and the very slow drop-off over time is far more significant than the recent adult content ban. If you choose not to believe me, you can go to Internet Archive / Wayback Machine and view post-per-day stats on https://www.tumblr.com/about https://www.tumblr.com/about going back a decade and see for yourself. I'm not going to reply to this further. I know the stats, I've seen the internal stats over many years, and this is a pointless thing to argue.
  - coldtea 7y ago
    >Well, "most users" implies a majority, which is simply not the case here, full stop. Well, 30% of the traffic leaving within the first couple of months of the ban, seems rather significant to me: "The blogging platform’s traffic has dropped by 30 percent since the December ban on all adult content (...) In December, Tumblr’s global traffic clocked in at 521 million, but it dwindled to a mere 370 million in February, according to The Verge". And moving onwards, it continues to look like a death spiral - a website that lost half or more of its users and is on a slope: https://www.statista.com/statistics/261925/unique-visitors-to-tumblrcom/ https://www.statista.com/statistics/261925/unique-visitors-t... >I'm not going to reply to this further. I know the stats, I've seen the internal stats over many years, and this is a pointless thing to argue. Sure, no reason to reply. I've posted the stats myself, people can draw their own conclusions...
    - evanelias 7y ago
      > it dwindled to a mere 370 million in February You are linking to incorrect stats. These numbers claim Tumblr has more MAUs than Twitter and Snapchat combined.
      - coldtea 7y ago
        Feel free to link to the "correct" ones...
        evanelias 7y ago
        Those stats are not publicly available. In general, for any company, at best they may or may not include the stats in quarterly earnings statements if the company is publicly traded. Although Tumblr has been owned by public companies since 2013, its corporate owners have not made these stats publicly available. Elsewhere in this subthread, I've linked to public info showing that in absolute content creation numbers, the impact of the adult content ban was far less than a majority of content, and also far smaller than the slow drop-off of users over the previous five years. Judging by your use of quotes around "correct", it seems you've already made up your mind anyway and I'm wasting my time discussing this. If you want to trust clearly incorrect numbers in The Verge over the person who built the company's relational storage tier, I suppose that's your prerogative.
        coldtea 7y ago
        >If you want to trust clearly incorrect numbers in The Verge over the person who built the company's relational storage tier, I suppose that's your prerogative. So, the Verge refers to data from two sources SimilarWeb and Statista, whereas your data are basically "trust me" and "absolute content creation numbers" (that is: not visitors). And on top, you say you've built Tumblr's infrastructure. I guess it's me and The Verge who is biased.
        evanelias 7y ago
        How would you feel if random people on HN repeatedly insisted your former employer was primarily a porn site, despite these people having no real evidence, and despite this going against years of your personal experience scaling and capacity-planning the site? Do you actually believe these numbers from SimilarWeb and Statista claiming Tumblr has more MAUs than Twitter and Snapchat combined, even after banning adult content (which you claim was a majority of Tumblr's usage)? Does that even remotely make any sense at all? You're outright accusing me of lying here. I find this insulting and do not wish to continue this discussion. Goodbye.
nesadi 7y ago
I've observed most discussions about Tumblr outside of HN. Most discussion revolves around how banning adult content from Tumblr removed any reason to go to Tumblr for many people. These are your average kinksters, artists, creative types who have been forced off the platform because it no longer caters to their interests. Not sure what any of this has to do with HN.
- evanelias 7y ago
  This subthread is discussing a misconception prevalent among HN users. But this same misconception may also be shared by other groups; no one is saying it's exclusive to HN. I believe your anecdotal experience in this area is true and accurate. But at the same time, the raw stats just don't reflect it being universally accurate across all users. A similar phenomenon has occurred regarding people saying they're quitting Facebook, deleting Uber, moving off GitHub, etc. In this specific case re: Tumblr, my guess would be that only a portion of them actually stopped using Tumblr at that specific time and for that reason. I suspect some of these people already stopped using Tumblr long ago, and meanwhile some others do continue using Tumblr despite previously saying otherwise. I'm just speculating though.
  - cf498 7y ago
    Do you base that on anything but a suspicion on your part? You are claiming this to be a misconception, can you actually back that up? I have no idea who those new content creators are supposed to be, like everyone else i just saw people with a decade or more of content producing behind them leaving. And not some of the people i knew but all of them. If there are great hidden communities that grew in the vacuum left, please share.
    - evanelias 7y ago
      I am basing my statements on knowledge from personal experience. As I said above, I'm a primary source with first-hand factual knowledge about this topic. (If you do a lot of capacity planning for a UGC / social network infrastructure over many years, you become intimately familiar with changes in growth rate over time...)

This is spot-on. HN has generally massively over-estimated the amount of porn on Tumblr, in terms of percentage of total content. I say this as a primary source with first-hand factual knowledge.

coldtea 7y ago

It doesn't have to be a large "percentage of total content". It just has to be the content that matters to most users...

evanelias 7y ago
Well, "most users" implies a majority, which is simply not the case here, full stop. Far from it, especially if you define a user as a person who has a Tumblr account and posts content on the site. If you consider a "user" to include non-monetizeable lurkers who were just there to view adult content, then saying "most users" is definitely still wrong, but perhaps slightly less so. Tumblr's peak was many years ago (2013 iirc), and the very slow drop-off over time is far more significant than the recent adult content ban. If you choose not to believe me, you can go to Internet Archive / Wayback Machine and view post-per-day stats on https://www.tumblr.com/about https://www.tumblr.com/about going back a decade and see for yourself. I'm not going to reply to this further. I know the stats, I've seen the internal stats over many years, and this is a pointless thing to argue.
- coldtea 7y ago
  >Well, "most users" implies a majority, which is simply not the case here, full stop. Well, 30% of the traffic leaving within the first couple of months of the ban, seems rather significant to me: "The blogging platform’s traffic has dropped by 30 percent since the December ban on all adult content (...) In December, Tumblr’s global traffic clocked in at 521 million, but it dwindled to a mere 370 million in February, according to The Verge". And moving onwards, it continues to look like a death spiral - a website that lost half or more of its users and is on a slope: https://www.statista.com/statistics/261925/unique-visitors-to-tumblrcom/ https://www.statista.com/statistics/261925/unique-visitors-t... >I'm not going to reply to this further. I know the stats, I've seen the internal stats over many years, and this is a pointless thing to argue. Sure, no reason to reply. I've posted the stats myself, people can draw their own conclusions...
  - evanelias 7y ago
    > it dwindled to a mere 370 million in February You are linking to incorrect stats. These numbers claim Tumblr has more MAUs than Twitter and Snapchat combined.
    - coldtea 7y ago
      Feel free to link to the "correct" ones...
      - evanelias 7y ago
        Those stats are not publicly available. In general, for any company, at best they may or may not include the stats in quarterly earnings statements if the company is publicly traded. Although Tumblr has been owned by public companies since 2013, its corporate owners have not made these stats publicly available. Elsewhere in this subthread, I've linked to public info showing that in absolute content creation numbers, the impact of the adult content ban was far less than a majority of content, and also far smaller than the slow drop-off of users over the previous five years. Judging by your use of quotes around "correct", it seems you've already made up your mind anyway and I'm wasting my time discussing this. If you want to trust clearly incorrect numbers in The Verge over the person who built the company's relational storage tier, I suppose that's your prerogative.
        coldtea 7y ago
        >If you want to trust clearly incorrect numbers in The Verge over the person who built the company's relational storage tier, I suppose that's your prerogative. So, the Verge refers to data from two sources SimilarWeb and Statista, whereas your data are basically "trust me" and "absolute content creation numbers" (that is: not visitors). And on top, you say you've built Tumblr's infrastructure. I guess it's me and The Verge who is biased.
        evanelias 7y ago
        How would you feel if random people on HN repeatedly insisted your former employer was primarily a porn site, despite these people having no real evidence, and despite this going against years of your personal experience scaling and capacity-planning the site? Do you actually believe these numbers from SimilarWeb and Statista claiming Tumblr has more MAUs than Twitter and Snapchat combined, even after banning adult content (which you claim was a majority of Tumblr's usage)? Does that even remotely make any sense at all? You're outright accusing me of lying here. I find this insulting and do not wish to continue this discussion. Goodbye.

nesadi 7y ago

I've observed most discussions about Tumblr outside of HN. Most discussion revolves around how banning adult content from Tumblr removed any reason to go to Tumblr for many people. These are your average kinksters, artists, creative types who have been forced off the platform because it no longer caters to their interests. Not sure what any of this has to do with HN.

evanelias 7y ago
This subthread is discussing a misconception prevalent among HN users. But this same misconception may also be shared by other groups; no one is saying it's exclusive to HN. I believe your anecdotal experience in this area is true and accurate. But at the same time, the raw stats just don't reflect it being universally accurate across all users. A similar phenomenon has occurred regarding people saying they're quitting Facebook, deleting Uber, moving off GitHub, etc. In this specific case re: Tumblr, my guess would be that only a portion of them actually stopped using Tumblr at that specific time and for that reason. I suspect some of these people already stopped using Tumblr long ago, and meanwhile some others do continue using Tumblr despite previously saying otherwise. I'm just speculating though.
- cf498 7y ago
  Do you base that on anything but a suspicion on your part? You are claiming this to be a misconception, can you actually back that up? I have no idea who those new content creators are supposed to be, like everyone else i just saw people with a decade or more of content producing behind them leaving. And not some of the people i knew but all of them. If there are great hidden communities that grew in the vacuum left, please share.
  - evanelias 7y ago
    I am basing my statements on knowledge from personal experience. As I said above, I'm a primary source with first-hand factual knowledge about this topic. (If you do a lot of capacity planning for a UGC / social network infrastructure over many years, you become intimately familiar with changes in growth rate over time...)