lies and damn lies, AO3 edition
There is a meme going around where people are posting the list of their most popular stories on AO3, comparing now vs one year ago, and - this is not that meaningful for me, because most of my stories aren't on AO3. But coincidentally, exactly a week ago I posted a new story to AO3 (only, though linked from my journals and a community journal) and I have been keeping daily statistics, because there have been recent discussions around fan-space about feedback patterns and rates on AO3, and I thought it would be interesting to see what the actual numbers said.
Some years back (2005 and 2006) I did a whole series of website statistics posts, mostly focusing on my own website vs livejournal hits, and feedback vs hits. (And how much feedback I got via email vs via lj-comment, which harks back to the sub-discussion on kudos about how the complaints about kudos taking the place of comments resemble the complaints about comments taking the place of email!) In general terms, I found that a story will get most of its hits (and by far most of its feedback) in the first week after posting (and most in just the first four days), and that stories in big fandoms get more hits, and more absolute feedback, but a lower ratio of feedback to hits.
So here are the raw numbers for one week of hits on The Forest of Dreams (~4500 words of Eagle fandom tentacle porn). I was not being extremely careful about counting EXACTLY at the same time each day, but they serve as a general snapshot:
day comments kudos lj adj new fb new hits hits tot fb ratio
1 1 14 2 1 16 129 129 16 0.124
2 4 24 3 2 13 196 325 29 0.089
3 6 29 4 3 7 70 395 36 0.091
4 7 29 4 3 1 36 431 37 0.086
5 7 29 4 3 0 31 462 37 0.080
6 8 30 4 3 2 29 491 39 0.079
7 8 32 4 3 2 20 511 41 0.080
comments = AO3 comments from others directly on my work (not discussion)
kudos = AO3 kudos (guest and signed)
lj = comments on my LJ or DW post, or on my pointer post at ninth_eagle.
adj = adjustment of (signed) kudos also left by a commenter.
new fb = that day's sum of comments at AO3 and journal sites and kudos, minus the adjustment.
new hits = that day's hits
hits = cumulative hits
tot fb = cumulative feedback
ratio = (tot fb)/hits = cumulative ratio of feedback per hit.
The feedback ratio is really only relevant at the first week, because the hits will continue to come (slowly) but the feedback at this point will be very, very sparse.
So, how does this stack up to the previous results? Part of the impetus for my post was a fandom friend who calculated a feedback ratio of 5-10% on her stories; I looked at three of my stories at the time and came up with very similar numbers. I got 17% for a very small fandom story (Wilby Wonderful, which I would have thought no smaller than Eagle fandom, but the absolute hits were way less), and 6% for an SGA story which is probably my most popular (in terms of absolute hits), and 4% for a longish smutless due South story. This story's ~8% ratio seems right on target.
It's hard to compare absolute hit numbers across fandoms. Other than challenge stories (Yuletide, Remix) I haven't posted anything new other than Eagle stories to AO3. But looking at my two most popular stories on AO3 as measured by total hits, amazingly my first Eagle story has an 8% or so feedback ratio (six weeks after posting!) and my most recent Yuletide story 10%. I have not taken out the signed kudos from people who also commented, so the real numbers are probably a little less, but this is still pretty darn good - and totally par for the course compared to previous stories posted pre-AO3.
Do kudos "steal" from comments? Hard to say (and I consider kudos yay! Feedback just like comments!) It's of course hard to quantify what a "comment" is - one of my AO3 comments is "Yay! Sweet!" But most of my AO3 comments are a whole sentence, if not several sentences. But - looking at my comments for Across the Great Divide (the longish smutless dS story mentioned above, written in 2006), most of them are REALLY LONG. I mean, paragraphs. Most of the AO3 comments are, as I said, a sentence or two. I wondered whether the kudos culture is contributing to shorter comments, but when I looked at a Yuletide story posted on AO3 before kudos existed, the comments are short there, as well. So maybe it's the AO3 (feedback) vs LJ/DW (conversation) culture that was mentioned in some of the kudos vs feedback posts.
However, I will say that the ratio of feedback via lj-comment (as opposed to email), in an early test, was 85%. (In another, it was 90%). The ratio of kudos to total feedback is about 80%. I have no idea what this means, but I note that kudos can be anonymous, while feedback on AO3 at least has to be with a sockpuppet name.
In conclusion, well, I can't really draw any conclusions. Other than that for a story posted now, in 2012, to AO3 and linked from LJ and DW, the feedback relative to hits is about the same as it was in 2006, for a story posted to my website and linked from LJ and DW.
Oh, yeah, and I donated some $$ to the OTW, which runs AO3. Imagine a cool linky image here.
Also posted on Dreamwidth where there are
comments.
Some years back (2005 and 2006) I did a whole series of website statistics posts, mostly focusing on my own website vs livejournal hits, and feedback vs hits. (And how much feedback I got via email vs via lj-comment, which harks back to the sub-discussion on kudos about how the complaints about kudos taking the place of comments resemble the complaints about comments taking the place of email!) In general terms, I found that a story will get most of its hits (and by far most of its feedback) in the first week after posting (and most in just the first four days), and that stories in big fandoms get more hits, and more absolute feedback, but a lower ratio of feedback to hits.
So here are the raw numbers for one week of hits on The Forest of Dreams (~4500 words of Eagle fandom tentacle porn). I was not being extremely careful about counting EXACTLY at the same time each day, but they serve as a general snapshot:
day comments kudos lj adj new fb new hits hits tot fb ratio
1 1 14 2 1 16 129 129 16 0.124
2 4 24 3 2 13 196 325 29 0.089
3 6 29 4 3 7 70 395 36 0.091
4 7 29 4 3 1 36 431 37 0.086
5 7 29 4 3 0 31 462 37 0.080
6 8 30 4 3 2 29 491 39 0.079
7 8 32 4 3 2 20 511 41 0.080
comments = AO3 comments from others directly on my work (not discussion)
kudos = AO3 kudos (guest and signed)
lj = comments on my LJ or DW post, or on my pointer post at ninth_eagle.
adj = adjustment of (signed) kudos also left by a commenter.
new fb = that day's sum of comments at AO3 and journal sites and kudos, minus the adjustment.
new hits = that day's hits
hits = cumulative hits
tot fb = cumulative feedback
ratio = (tot fb)/hits = cumulative ratio of feedback per hit.
The feedback ratio is really only relevant at the first week, because the hits will continue to come (slowly) but the feedback at this point will be very, very sparse.
So, how does this stack up to the previous results? Part of the impetus for my post was a fandom friend who calculated a feedback ratio of 5-10% on her stories; I looked at three of my stories at the time and came up with very similar numbers. I got 17% for a very small fandom story (Wilby Wonderful, which I would have thought no smaller than Eagle fandom, but the absolute hits were way less), and 6% for an SGA story which is probably my most popular (in terms of absolute hits), and 4% for a longish smutless due South story. This story's ~8% ratio seems right on target.
It's hard to compare absolute hit numbers across fandoms. Other than challenge stories (Yuletide, Remix) I haven't posted anything new other than Eagle stories to AO3. But looking at my two most popular stories on AO3 as measured by total hits, amazingly my first Eagle story has an 8% or so feedback ratio (six weeks after posting!) and my most recent Yuletide story 10%. I have not taken out the signed kudos from people who also commented, so the real numbers are probably a little less, but this is still pretty darn good - and totally par for the course compared to previous stories posted pre-AO3.
Do kudos "steal" from comments? Hard to say (and I consider kudos yay! Feedback just like comments!) It's of course hard to quantify what a "comment" is - one of my AO3 comments is "Yay! Sweet!" But most of my AO3 comments are a whole sentence, if not several sentences. But - looking at my comments for Across the Great Divide (the longish smutless dS story mentioned above, written in 2006), most of them are REALLY LONG. I mean, paragraphs. Most of the AO3 comments are, as I said, a sentence or two. I wondered whether the kudos culture is contributing to shorter comments, but when I looked at a Yuletide story posted on AO3 before kudos existed, the comments are short there, as well. So maybe it's the AO3 (feedback) vs LJ/DW (conversation) culture that was mentioned in some of the kudos vs feedback posts.
However, I will say that the ratio of feedback via lj-comment (as opposed to email), in an early test, was 85%. (In another, it was 90%). The ratio of kudos to total feedback is about 80%. I have no idea what this means, but I note that kudos can be anonymous, while feedback on AO3 at least has to be with a sockpuppet name.
In conclusion, well, I can't really draw any conclusions. Other than that for a story posted now, in 2012, to AO3 and linked from LJ and DW, the feedback relative to hits is about the same as it was in 2006, for a story posted to my website and linked from LJ and DW.
Oh, yeah, and I donated some $$ to the OTW, which runs AO3. Imagine a cool linky image here.
Also posted on Dreamwidth where there are
comments.