Wednesday, February 28, 2007

Percent who tag

Some bloggers* has talked about my statement in When tags work and when they don't, that:
"Tags work best when they're about memory, so tagging makes the most sense when you have a lot of something to remember. On LibraryThing, members with under 50 books seldom tag, but users with 200 or more usually do.** When you get right down to it, few of us need to remember 200 books on Amazon. For most of us, the "wishlist" feature is good enough. We don't need to sub-segment out the "anthropology" books."
Here's some data on that issue. I compared the number of books a LibraryThing member has with whether they tag or not. The later is defined as having at least one tag, so it over-represents taggers. But the trend is clear. The more you have to keep track of, the more you tag.



(click to enlarge)

*Notably, Thomas Vanderwal, who coined "folksonomy." Vanderwal is giant. He makes some criticisms of my post that I agree with—I wrote pretty rapidly and off-the-cuff. And some I don't. I have not been watching Amazon's tagging month by month as Thomas has, but his points don't change my mind. Even with half the time—and remembering that LibraryThing was dead-obscure for much of it's life—and any other caveats and nuances one applies, Amazon's tagging experiment hasn't worked out. If it worked, they'd be swimming in high-quality tags. They aren't, and they've been distracting customers and burning up valuable screen space to acquire a web of meaning so flimsy as to be largely useless for its ends. Anyway, I hope to get a reply out soon.
**You'll note my qualifiers are a bit off. But my impressions match the facts better insofar as the numbers above overplay taggers. In theory, I could set a bar—X% of books are tagged. But that would miss some people who catalog first, tag second. I've seen that a lot—although I'd go mad if I delayed it like that! Those people are taggers too.

Post footnotem: Anyone have an explanation for why tagging dips at 200-250? People who hit the free-200 wall, get frustrated and leave before tagging?

18 Comments:

Blogger Zac said...

That's one o' them "outliers" ... you could as easily ask why it dips at 1600.

I think another interesting graph would be "nbr books" vs "% of books tagged".

3/01/2007 12:46 AM  
Blogger Tim said...

>That's one o' them "outliers" ... you could as easily ask why it dips at 1600.

No. I don't think so. The number of users at 200-250 is enormous. The number at 1,600 is smaller. At one point you could estimate the latter by looking at the top 100 libraries, but that now starts above 2,000.

3/01/2007 12:51 AM  
Blogger Zac said...

Then I suspect it is the free 200 wall after all. The catalog first, tag later users who figure there's no point in tagging until after they buy the account, which they haven't done yet.

You could support that by generating the same graph, but with separate bars for the paid vs unpaid accounts.

3/01/2007 2:45 AM  
Blogger Amanda Ellis said...

Maybe they're busy stacking up books to catalog on mass.

3/01/2007 5:53 AM  
Blogger gabriel said...

Hmmm. If it is indeed the "free 200 wall", I'd expect those users with precisely 200 books to have a substantially lower tag rate than either those approaching 200 or slightly above 200. That would account for the dip.

But I'd suggest that a different dynamic may be at play here. Users over 200 don't have a compunction about buying the account, and are apt to see their ~220 books as less significant than someone with 180 books and are careful with their money.

3/01/2007 6:43 AM  
Blogger nicole said...

I think it has to do with the free 200 wall also, but for a somewhat different reason. A user with 150 books who knows he's going to buy in isn't tagging with 150 books in mind - he's tagging with the thought of putting money into this and definitely having over 200 books. A user who hasn't made the mental commitment to buy a membership is tagging in the same mindset from the point of first entry to the point of two hundredth entry - that is, he's tagging (or not) counting on only having 200 books ever.

I guess part of what I'm saying is that what appears to be a "dip" is actually more like what you *should* see at a number like 200, whereas what you see at a number like 150 is affected by people who know they'll eventually have 500 or 100 books catalogued.

3/01/2007 7:59 AM  
Anonymous comfypants said...

It could be that rather than people with few cataloged books not tagging, it's people with no tags not cataloging. That is, they stop at a low number because, not having taken advantage of what LibraryThing can do for them, they don't have as strong a reason to continue.

3/01/2007 11:40 AM  
Anonymous sunny said...

Would it help to break the 200-250 down to steps of 10? Or look at 201-250 (excluding the accounts that stopped at 200).

Also, as zac said: distinguishing between paid and unpaid might show some explanation.

3/01/2007 6:07 PM  
Blogger Andrew said...

zac: Number of books versus average number of tags per book would be interesting, too.

3/01/2007 7:23 PM  
Blogger James said...

Post footnotem: Anyone have an explanation for why tagging dips at 200-250? People who hit the free-200 wall, get frustrated and leave before tagging?

I don't think it can be that, as we are seeing the dip in people who have actually paid for 200 books or more. Unless there are a huge number of users exactly on the 200 book boundary stuffing up the stats?
Are you allowed to put your 200th book in for free or not? Maybe the graph would look different if the scale was offset by 1?

My Prediction as to why the dip, it is all those catalogue first and tag later crowd. They are only just getting getting started on there way to 1000 books or something of that nature. I'm guessing most people who pay for an account and actively use it have a lot more that 200 books.

Personally it amazes me that people would catalogue 100 (or much much more) without even having a go playing with the tagging! Surely you would think to try adding a tag just to see what it does?

3/01/2007 10:14 PM  
Anonymous _Zoe_ said...

I think that dip shows that people who don't care about money are most likely not to care about tagging. Those are the people who are willing to pay $10 or $25 for a service that they're really not making the best use of. If you have 210 books, is it really worth it to pay $10 to catalogue those last 10? People who spend money like that may not value what they're purchasing. On the other hand, people with 200 or fewer books may really care about their catalogues and their books but just not have the money for a paid account.

3/02/2007 1:09 PM  
Anonymous BTRIPP said...

I don't know ... if I didn't tag for "library order", I probably wouldn't tag at all.

I find the concept of "content tagging" very daunting. There are, of course, "top of head" tags where glancing at a book I'll say "SciFi" or "War" or "HighSchool", and maybe some folks are willing to go with that ... and, frankly, were I to "have to" go back through my entire library to put content tags on things this might be the level I'd initially have to aim for.

However, (with my ever-present OCD), this is not enough, so I would need (on some deep psychological level) to spend a substantial time with each and every book to wring out all the "essential things" about it, which would probably take years and years of undivided attention to "fully tag" my whole library ... and to what end? Do I need to know what books are (to pull an example out of the aether) set in Boston? Would I ever need to sort books that way?

For me, that's the problem with tagging ... where is the point where it's useful for extracting information but not more detail than I'll ever need? So far I'm happy to just have the books in my L.T. collection show up in the order they're on the shelves of my physical library!

3/03/2007 6:45 AM  
Blogger "As You Know" Bob said...

Well, to generalize from my own LT experience:

The people who have broken through the "200-book" limit are relatively early in their LT experience, and are getting their books entered first, and are planning to go back and play with tagging at some later time.

"200" is a relatively hard ceiling; the 200-250 level is therefore more likely to be a very transient, temporary position.

So I would expect that people have a real financial incentive to linger at the 'below-200' level and play with (and tag) their collection. Whereas somebody at 201 books is more likely to be "getting the stuff entered first".

3/03/2007 1:02 PM  
Blogger wunder said...

To avoid the "catalogue first, tag later" problem, omit new signups. Restrict it to people who've been around for three months or more, for example.

3/05/2007 12:33 PM  
Anonymous MyopicBookworm said...

I agree with AsYouKnow_Bob: I bet most people in the 200-250 range are those who have just moved from free to paid, and are focusing on adding first, tagging later. I'm now tagging as I go, but when (in the far distant future) I actually get most of my books catalogued, and can't afford the money or space to keep buying more, then I'll probably feed my LibraryThing addiction by devising a more and more elaborate tagging system.

3/09/2007 12:00 PM  
Blogger Rachael said...

I know I'm coming late to the table, but I had to comment. Here's my theory, convoluted though it may be: the people with less than 200 books are more likey to be persons who are very young. I think this is indicated by the size of their library. They tag because they know what tags are and what they accomplish. The person that enters 200 books and no more may very well have a much larger library but be older and unfamiliar with tagging (like my mother) and not understand the full capabilities of Librarything. They get their books cataloged but then lose some interest because they're not quite sure where to go from there. Unwilling to pay for membership, they simply leave the 200 books cataloged and then quit.

3/15/2007 8:42 PM  
Anonymous johnascott said...

I also think a.y.k. Bob makes a good point. Maybe this can be tested by comparing tagging rates between folks in the 150 - 200 range who have paid (and for whom the 150 - 200 is presumably a transient state) and those who haven't paid?

3/25/2007 3:28 PM  
Anonymous Kit said...

I think the issue has more to do with the difference between people who consider themselves collectors/librarians, and people who just have books. IMO people who haven't had cataloging experience aren't as likely to tag. My bookcases are designated by fiction, non-fiction, childrens. Then within that the books on each shelf are arranged by topic/genre. That is going to make my tagging much easier, because mentally each book already has tags. Plus, I have library cataloging experience, so many of my tags will come straight from the cataloging info.

3/25/2007 8:56 PM  

Post a Comment

<< Home