Skip to content

Tagging in workflow context

The personal detour

I’m back from giving the closing talk at SOLINET’s annual membership meeting, where I was well-pampered by all involved. I also rented a Prius for the trip, and it was quite difficult to return this lovely car to Avis. I had wondered how I would like driving a Prius, and the answer is “OMG ponies!” Smooth ride, the joy of watching myself save fuel, and pride of (temporary) ownership of a green vehicle… ‘sall good.

I didn’t sleep well for two nights running (no fault to anyone except my over-active brain working on issues completely unrelated to SOLINET, speeches, or cars), so during my talk I felt under my game. I can always feel the difference between “they liked my talk” and “I shot them over the moon with numinous insights.” I’ll make a point of sleeping better before I teach “Writing for the Web” at TBLC next month.

The when-ness of tagging

Now I make a sharp right turn to discuss tagging in workflow context. Over on Thingology, Tim Spalding discusses user tagging of Godless, Ann Coulter’s latest screed book, pointing out that on Amazon the shouting match is unrelated to book ownership:

But while, on LibraryThing, where you have to have a book to tag it, Godless has a fairly unremarkable tag cloud, touching on its subject matter and point of view, on Amazon, the tagging has devolved into a shouting match.

For some time I’ve been pondering tagging in the context of a user’s workflow. Tagging in library catalogs hasn’t worked yet for a number of reasons, such as these rather obvious points:

  • John Blyberg has noted that without critical mass, tagging is useless. I’d go farther and say without critical mass, tagging could backfire, because only the most determined cranks and pranksters might actually use it. A local library catalog is not beefy enough to build critical mass on its own; I don’t know how big or how heavily-used a catalog needs to be, but “a lot” is my guess. (Then there is the issue with the silo-like design of most library software, which keeps social data imprisoned behind proprietary walls.) That is yet another reason I like “LibraryThing for Libraries“: it’s an enrichment service to salt a catalog with an initial mass of high-quality tags built by passionate readers (and also provides that spookily-marvelous if-you-liked-this functionality).
  • Some systems that claim to offer tagging make it so high-pain to tag that it works against adoption. I am thinking of the system where to merely SEE the tags a user must log in, and where tags are only searchable in “Advanced Search.” (Carl Grant, if you’re reading this, I owe you a citation on people-don’t-use-advanced-search… you have been very patient.)
  • Also, on several occasions I have observed conversations about tagging between vendors and customers where the first words out of a customer’s mouth are “How can I control tagging?” and the vendor then responds in kind. If your primary objective is to “control” tagging, rather than make it work (that is, at minimum, to encourage users to provide quality tags), then the system design, to borrow youthful jargon, will be a FAIL.

But I have also pondered tagging in workflow context and feel this has not received adequate discussion. I’m guessing (based on Tim’s comments) that Librarything users are predominantly tagging when they add books or when they return to their collection for maintenance/grooming activities, such as cleaning up entries, fiddling with their default display, or examining the community discussion around books. Tim is also suggesting that on Amazon tagging appears to be less related to activities related to the workflow of book acquisition and ownership.

So I again mull over the library catalog and tagging workflow. Most catalogs are designed to help users find books or book-like items — known items, or items found through discovery. (Well, that is the claim, anyway.) You don’t return books through a library catalog (at least not yet). So when would tagging happen?

My guess is the best tagging would happen when the users returns the catalog to find more items. I say this because in some respects, a library catalog appears to be remarkably similar to Netflix in workflow, where I (again, out on this limb!) presume user reviewing (similar to tagging?) happens when a user logs in to refresh his or her queue with yummy new titles or simply get a reminder of what’s in the queue (in my family’s case, this happens after we receive some bizarre movie that sorta-looked-good that stealthily crept up to be #1 in one of our queues).

If I’m not going to tag when I find a book (why would I, if I haven’t read it, Amazon notwithstanding), and I’m not going to tag when I check out a book (an unrelated physical activity), and I’m not going to tag after I read a book (because that would mean the sole reason I’m returning to the catalog is to tag an item, which feels low-gain), and I’m not going to tag when I return a book (can you see me at the circ desk, reciting tags I want added to an item — or perhaps shouting tags into a book drop? Or I guess I could write them on a p-slip)…

Seems to me that tagging workflow in a catalog should be “gamed” so that the next time I visit the catalog to find something, the catalog entices me to tag. That would also be when I’m motivated to tag the book in a way that describes it well for my own bibliographic reuse, and also for others. (It could lead to opinion-tagging, though maybe that is always inevitable.)

Then again, what if at the beginning of a new discovery session the catalog recommended books? Prompted me to add reviews? Suggested I queue items? But I get ahead of myself…

All I’m really saying is that the very primitive tagging workflows I’ve seen so far in library catalogs aren’t designed to encourage tagging. (I am not referring at all to LibraryThing for Libraries, which at this point is a one-way enrichment service.) In fact, I don’t see much attention to tagging workflow, period. It feels very random and first-gen — a tacked-on service to allow a vendor to say “Yes, we offer tagging.” If you care at all about engaging users in catalogs and building user-contributed data, or for that matter leveraging social data period, that is simply not good enough.

Thoughts on tagging? Do I have this all wrong, or is there a nubbin of sensicalness here? Have I missed or misinterpreted/misrepresented some tagging behavior?

Posted on this day, other years: