Extending the Reach of Tagging

Share via Twitter Share via Facebook Share via Linkedin Share via Reddit

Given my recent experiences – nearly all positive – with metadata enabling services such as del.icio.us and Flickr, it’s no surprise that I’m seeking ways to expand metadata hooks into new areas of my overall information management scheme.

Joi’s recent post on getting “taggy” illustrates the appeal nicely, I think. I’ve called this the “network motivation” in the past (see here, or here), in the sense that the network amplifies the returns of my own tagging efforts [1] and thus serves as a powerful secondary motivation for capturing metadata and data in a social, tagged context.

I was reminded of the tangible benefits a few days ago when I had to call a corporate travel agency to make arrangements for a client engagement. The agency needed to look up my profile, and I quite naturally didn’t have an employee number because I’m not an employee. That required her to “go in” on my name. Normally, this is probably not much of a challenge, but the system apparently needed an exact match, so it needed to know whether or not I had a middle initial in my profile. Well, I do have one, but whether I had one in there was a good question. To make a long story short, we had to query the database a few times before pulling me up. Now, obviously this is a simple case of poor system/database design, but the rep’s notion of “going in” on a different field was the takeaway from this.

We’re all implicitly familiar with this concept – we do it every day when we use Google. As Dare Obsanjo points out, we do it with iTunes. But by not explicitly supporting extended tagging, our “knowledge worker” (for lack of a better term) applications – I’m speaking of PIM and Office Productivity tools, primarily – do little to assist us in “going in” on different terms later during the often difficult task of retrieval. Hence the need for high overhead solutions such as GDS and full text search, and longer term a complete overhaul of the Windows file system in WinFS.

Now the pushback from Microsoft, Open Office, and others is likely to be that they already support metadata storage, and that it’s simply underutilized. And actually, I agree with that. Microsoft Office, for example, allows users to enter fields such as Title, Subject, Author, Company, Category, Keywords, and more. Despite that ability, the metadata is massively ignored, and more likely to receive attention as a security risk (even to the extent that it requires “cleaning“) than a feature enabling better productivity. So what gives?

Well, it’s clear that a few things are at work here. First, the metadata support embedded in these applications is traditionally buried (File: Properties – doesn’t even get its own window) where users will never find it, unless they stumble upon it by accident. Second, even were they to know where to put the metadata and feel like keeping it updated, the downstream benefits of this additional effort are not made apparent. So why bother? Third, vendors, customers and users are still figuring out how to use this metadata. Flickr and its brethren are but recent discoveries, and have converted a relative minority of users on the benefits of social, collaborative metadata. But I think on the latter count, Pandora’s box has been opened, and the benefits are clear to a large and growing population of early adopters, and that metadata/tagging in on a hockey stick style growth curve (which maybe explains some of the problems I’ve had with the web services lately).

I for one am hoping that ISVs and open source projects alike are watching this trend closely, and steal a page from the current metadata flavors of the month. Where might this be advantageous?

1. Working on an expense spreadsheet, I have a tagging mechanism readily exposed in the interface, which allows me to apply tags like “billable”, “clientname”, “2004”, and “travel” for later retrieval.

2. Within my PIM client, I’m able to tag an email from an important client with “clientname”, “projectname”, “todo”, “retain”, and a mix of keywords so that I can sort easily on multiple views later. Boris articulated this nicely in a post from yesterday.

Ideally this data would be centrally accessible on the desktop and (optionally) network aware, but an incremental adoption within the applications themselves would be a great start. And more than likely this type of improvement would be a relatively trivial extension given that as mentioned rudimentary metadata facilities are already present in many cases. First step is to expose the tagging underpinnings you’ve built-in in the simplest fashion possible (toolbar/sidebar, most likely).

You listening, Apple, Gnome, KDE, Lotus, Microsoft, or Open Office? Because this simply screams potential differentiator.

[1] Take a case of simple bookmarking. I can do one of the following:

a.) bookmark an item in my browser
b.) bookmark an item in del.icio.us (or Furl, Jots, or Simpy)

If I choose (a), the value of that bookmark is marginal, in that it’s simply being stored for later retrieval. If I choose (b), however, I open the door to new possibilities – I can find related links, I can find people who find related links, I can research more general topics, I show people what I’m linking to, etc. The network thus inflates the value of that simple bookmark, with a minimum of additional effort.

No Comments

Leave a Reply

Your email address will not be published. Required fields are marked *