Live data from Hacker News

Happy 10th birthday to us

redditblog.com

51–60 of 108 posts

Re: Happy 10th birthday to us

#51
post #47

Earlier quoted context omitted.

That seems very unusual to have a large uptick in engagement without any corresponding increase in traffic. I'm not sure I understand how the Digg announcement would make existing Reddit users more active.

I imagine the existing users were both Digg and Reddit users, and started using Reddit more when they stopped using Digg.

Which one might expect would lead to more page views which jedberg said didn't happen. They didn't use the site more, they just became more engaged.

Re: Happy 10th birthday to us

#52
post #11
post #4

Is Reddit still run to be profitable? It always seems like it fits more with a wikipedia type model of donations etc.

I'm more interested in how imgur survives since they serve up far more bandwidth that reddit does and their own advertising and community efforts are pretty young relatively, not to mention hotlinks bypass that entirely.

Imgur uses the free/extremely cheap CDN Cloudflare. Same with 4chan. The vast majority of Imgur's content gets served at the CDN layer so Imgur doesn't have to pay for that bandwidth.

moot (of 4chan) said a while back that Cloudflare is one of the main reasons 4chan is still around these days.

Re: Happy 10th birthday to us

#53
post #11

Earlier quoted context omitted.

I'm more interested in how imgur survives since they serve up far more bandwidth that reddit does and their own advertising and community efforts are pretty young relatively, not to mention hotlinks bypass that entirely.

Imgur uses the free/extremely cheap CDN Cloudflare. Same with 4chan. The vast majority of Imgur's content gets served at the CDN layer so Imgur doesn't have to pay for that bandwidth. moot (of 4chan) said a while back that Cloudflare is one of the main reasons 4chan is still around these days.

CloudFlare told me that using them to host images/media isn't allowed. That they do the CDN bits for websites, but not to do excessive non-webpage stuff. Their ToS says something like that.

Maybe that changes for a few K a month?

Re: Happy 10th birthday to us

#54
post #45

Earlier quoted context omitted.

Traffic didn't change much, but user behavior definitely did: http://minimaxir.com/img/reddit-statistics/all_score_reddit.... http://minimaxir.com/img/reddit-statistics/all_comments_redd... Source: http://minimaxir.com/2014/12/reddit-statistics/

Yes there was a definite uptick in commenting and submissions, but the overall traffic change (uniques and page views) wasn't huge.

Actually, it looks like both doubled in six months:

    Jul 2010,"7,275,451","279,891,841"
    Jan 2011,"12,284,583","590,531,477"
And then a couple months later we realized we were undercounting and it jumped again. :)

    Feb 2011,"12,654,137","594,987,855"
    Mar 2011,"16,925,815","1,133,827,794"

Re: Happy 10th birthday to us

#55
post #31
post #27

Earlier quoted context omitted.

> Reddit, who was struggling at the time. This is totally incorrect, BTW. Our traffic was already double Digg's at that point. Their v4 failure caused a noticeable but small bump in traffic.

That's actually really surprising. I was part of the Digg exodus and I'd barely heard about Reddit at the time of the v4 debacle, while Digg was a pretty big brand in tech circles. Was it just better marketing while Reddit actually had a larger-but-quieter userbase at the time?

As a counter-anecdote, by 2010 I had been a reddit user for 4 years and I had never commented nor had had an account on digg, which felt more like the flashier version of an online phpbb forum.

I had first heard about reddit in the summer of 2005 while holding my first ever job as a programmer doing python work, which I would consider as a "tech circle". I remember that I was a little bummed when the front page first got rid of technical-related content and forced me to go to programming.reddit.com instead. That and the fact that people like Paul Graham were still being judged based on their programming merit not on any crazy start-upy valuations certainly made reddit a very techie community in the beginning, at least from my point of view. And I almost forgot, many people were actually pretty upset when reddit switched from Lisp to Python (in December 2005, I even wrote on my blog about it), because python felt "less pure". Funny times.

Re: Happy 10th birthday to us

#56
post #54
post #45

Earlier quoted context omitted.

Yes there was a definite uptick in commenting and submissions, but the overall traffic change (uniques and page views) wasn't huge.

Actually, it looks like both doubled in six months: Jul 2010,"7,275,451","279,891,841" Jan 2011,"12,284,583","590,531,477" And then a couple months later we realized we were undercounting and it jumped again. :) Feb 2011,"12,654,137","594,987,855" Mar 2011,"16,925,815","1,133,827,794"

Yes, but that was our traffic trajectory before Digg v4 too, wasn't it?

Re: Happy 10th birthday to us

#57
post #50
post #27

Earlier quoted context omitted.

> Reddit, who was struggling at the time. This is totally incorrect, BTW. Our traffic was already double Digg's at that point. Their v4 failure caused a noticeable but small bump in traffic.

I'm not sure we were quite double their size. In July 2010, Digg claimed to have 200M pageviews[0] and we had 279M[1]. So I think we were more like 40% bigger in terms of pageviews, which Chris's post from the time[2] seems to confirm. Though we were undercounting a bit[3], so it was probably more like 55% or 60%. And when it comes to uniques, my memory's pretty foggy, so please correct me if I'm wrong, but I think D…

> And when it comes to uniques, my memory's pretty foggy, so please correct me if I'm wrong, but I think Digg actually had more than we did at that point.

My recollection was that when we visited them shortly after, we were shocked to learn just how much bigger we were. But we might both be right. It was a while ago.

Re: Happy 10th birthday to us

#58
post #51

Earlier quoted context omitted.

I imagine the existing users were both Digg and Reddit users, and started using Reddit more when they stopped using Digg.

Which one might expect would lead to more page views which jedberg said didn't happen. They didn't use the site more, they just became more engaged.

Keep in mind that commenting and voting don't count as page views since it's all AJAX. If you make no comments, one, or 50 on the same page, it's still one page view.

In fact, that was a main reason cited be people at the time as to why they liked reddit over other options -- because we didn't artificially inflate page views by forcing reloads.

Re: Happy 10th birthday to us

#59
post #56
post #54

Earlier quoted context omitted.

Actually, it looks like both doubled in six months: Jul 2010,"7,275,451","279,891,841" Jan 2011,"12,284,583","590,531,477" And then a couple months later we realized we were undercounting and it jumped again. :) Feb 2011,"12,654,137","594,987,855" Mar 2011,"16,925,815","1,133,827,794"

Yes, but that was our traffic trajectory before Digg v4 too, wasn't it?

IIRC we were doubling every 8-12 months, and Digg sped that up a bit for us that year. It wasn't like a make-or-break moment, but we certainly felt it. :)

Fortunately you had just turned up a bunch of new servers just days before.

Re: Happy 10th birthday to us

#60
post #47
post #45

Earlier quoted context omitted.

Yes there was a definite uptick in commenting and submissions, but the overall traffic change (uniques and page views) wasn't huge.

That seems very unusual to have a large uptick in engagement without any corresponding increase in traffic. I'm not sure I understand how the Digg announcement would make existing Reddit users more active.

It depends on how "page views" is defined, but I can imagine a case where the same number of pages were viewed, but the interactions on that page increased as people began commenting more on reddit than on digg. Reddit doesn't refresh pages when you comment, so if they didn't count the xhr requests as "page views", you could have more active, engaged users without significantly increasing the actual number of pages viewed.
Post reply on HN