Discovering versus teaching principles of social information management

In response to Josh Catone’s observation that has failed to go mainstream, Richard Ziade offers three hypotheses:

  1. Nobody really needs a way to centrally store their bookmarks
  2. Most people don’t understand what does
  3. People don’t feel compelled to share with others

The winning explanation, I am sure, is #2. Nobody understands what does. I am constantly explaining the nature and value of its social information management capabilities. Just this week, in various meetings on Microsoft’s Redmond campus, I found myself reiterating four of my major uses of

1. Answering a question with an URL.

I’m often asked questions like “What have you written about how to do screencasting?” I answer with an URL:

This not only wildly efficient, it’s future-proofed. If I hand you that URL today, then later add new items to the list, you’ll pick them up if you visit the URL in the future.

2. as a database.

The URL shown above is an example of the pattern I discussed here. It’s actually a query: select all bookmarks where one tag is screencasting and another tag is howto. If you understand that such queries are possible, judicious assignment of tags becomes a data management discipline.

3. Collaborative list curation.

As discussed here:

Recently I began keeping track of interesting public data sources using the tag judell/publicdata, and invited others to do the same using their own accounts. That method sets up an interesting pattern of collaboration whereby all contributions flow up to the global bucket, tag/publicdata, but individual contributors can curate subsets of that collection according to their own interests.

It’s a powerful pattern for loosely-coupled collaborative list-making.

4. Feedback monitoring.

When I’m visiting an URL, I often use my citations bookmarklet (available here) to see who has bookmarked the URL, which quotation and tags were used to describe it, and what the history of attention to that URL has been over time.

Is it’ fault that, even in the geek subculture where the service is mainly used, so few people seem to discover and exploit these patterns? I wonder about this all the time, and not just with respect to True, all of our information management tools could do a better job making features more easily discoverable. But to grok the patterns and apply the strategies I’m talking about, it’s not enough to know that features exist. You need to develop a sense of how those features can be used in support of certain principles of personal and social information management. It would be great if we could create software that naturally leads us to the discovery of those principles. But that’s a tall order. While we’re waiting, I think we should admit that these principles ought to be part of what you learn in order to become a digitally literate 21st-century citizen.

Posted in .

39 thoughts on “Discovering versus teaching principles of social information management

  1. Jon

    Wonderful post. I must say that of all the apps in the web 2.0 world, is the one I use the most, constantly. I’ve struggled to explain to people what makes it great. I think I’ll just point to your post instead.

    I often wonder if Yahoo did the right thing by not doing too much with, or were they criminally negligent by not realizing what an asset they have?

  2. This fall I required my graduate students (all K12 teachers) to obtain a delicious account and use it in class for sharing resources. The first couple of weeks, it was “Why do I have to….”. and then it became “Why did you not tell us about this sooner!” In their end of semester reflections, most noted that delicious was the most treasured find of the course.

  3. You need to add a restriction for at least points #3 and #4: you not only need to understand the benefits, but you need to have or build your (social) network for sharing the feature. This is so uncommon (no, you can’t just give your gmail contact list…) that most people will stop here.

  4. Yes. We’ve seen a similar effect with CoScripter. When we interviewed users, many asked if they could send a script to someone else — they never even considered copying and pasting the URL for their script into an email! That’s an even more basic web operation than your example of sending a URL to a tag collection on In order to accomplish your #1 and #2 above, people need to both realize that they can do that database query, and that they can refer to the results using a stable URL. I’m coming to believe that both those operations are still way beyond the capabilities of mainstream web users.

  5. Potentially related is how we use comments on posts just like these.

    Because Jon always provides useful information, and interesting people gather around his posts, I make it a point to read comments to his (and other’s) posts and follow the links behind people’s names. The blog post and associated comments themselves become a

    So, today I found Tessa’s link to CoScripter, a technology that is not only interesting in its own right but also highly related in a way to social information management. I promptly added it to my own bookmarks. Maybe CoScripter could automate this process?

    Disqus (I have no affiliation with them) seems to be positioned well to innovate in this area with comments.

    But I still sense a lack of a comprehensive system for me to manage all the sources of new links, connected with my various conversations on the web, and combine that with the automated discovery of new information relevant to me. solves one problem very well but there is still a huge opportunity here.

  6. “you not only need to understand the benefits, but you need to have or build your (social) network for sharing the feature.”

    Agreed. And since no single social bookmarking service will (or should) rule them all, this leads again back to the idea of federating, e.g., and connotea and furl and ….

  7. “I’m coming to believe that both those operations are still way beyond the capabilities of mainstream web users.”

    Clearly. Hence my point that, in addition to striving to make these things naturally discoverable, there are aspects of digital literacy that it’s high time we start to simply teach people, in grade school, along with the 3 Rs.

  8. I can understand the points here, but for a person who’s not interested in the process (ie, the vast majority of web users), try comparing point-by-point with Google or Yahoo search. I can see more effort required to use, for dubious additional benefit.

    1. Answering a question with an URL. “Just Google screencasting + howto.”

    2. as a database. Tags in a URL are the same as keywords in a search field.

    3. Collaborative list curation. I can’t think of a Google equivalent for this, but it sounds like a significant effort, and I can’t think of much incremental benefit for a mainstream web user.

    4. Feedback monitoring. If a user even cares about this, they can get it from the Google toolbar and its site ranking.

    I have to conclude that a web search that requires almost no effort and gives good results will satisfice most users, compared to building a presence at some effort in order to return slightly better results.

  9. “I can’t think of much incremental benefit for a mainstream web user.”

    To the extent that being a mainstream web user today means not having an interest in gathering and curating a persistent collection of online resources — both your own and others’ — that’s undoubtedly true.

    I suspect that at some point, mainstream web users will gather and curate as a matter of course. Especially as their own personal information management migrates into the cloud, where it will start to become evident that personal information management can work hand-in-hand with social information management.

  10. Wonderful post. I must say that of all the apps in the web 2.0 world, is the one I use the most, constantly. I’ve struggled to explain to people what makes it great. I think I’ll just point to your post instead.

  11. Hi Jon,
    thanks a lot for your really useful articles on tagging, bookmarking et al. Having read some of your thoughts I now have an idea how to explain to others, let’s say, the differences of “folders” and “tags” or what kind of revolution web2.0 usage patterns mean…

    Best regards,

  12. Hi Jon:

    As someone who cares a lot about URLs and how they can be used I agree that has much latent value. The problem is they haven’t created the killer interface for it yet. What will that killer interface be? I don’t know yet, but I’ll know it when I see it.

    But seriously, the problem I find with is several fold:

    1.)’ search is just too damn slow.

    Whenever I consider wanting to search, I have this subconscious aversion and experience an overwelming feeling of mental fatigue. Make it as fast as a Google search and this problem will go away.

    2.) doesn’t do any stemming or help deal with mispellings and/or synonyms.

    Sometimes I tag things with singular and sometimes plural, but they really are part of the same set. It takes a very structured mind with lots of discipline to be 100% consistent, and I unfortunately don’t have either of those attributes. Sometimes I use different words that mean the same, and certainly that happens across multiple people. Although the URLs are different, could offer acknowledgement of potentially related tags.

    3.) hides its tag management tools.

    Okay, “hide” might be too strong but could really make it a lot easier and more obvious how to go about merging, pruning, and otherwise managing tags. It could make suggestions, and/or at least off the tools in the context of tag/search results listings.

    4.)’ UI doesn’t offer drill-down of categories making it much, much less useful.

    For example, let’s say I tag something with “urls rewriting advocacy” and let’s assume that I’ve tagged lots and lots of articles with “urls rewriting” and lots of others with “urls advocacy.” Actually, you don’t have to assume because I have.

    Now if I go and search “urls” on from my favorites: will show me that I have 55 tagged with “urls+advocacy” and 38 tagged with “urls+rewriting” (along with a bunch of others that are also tagged with “urls.”)
    However, if I click the links for either “advocacy” or “rewriting” I don’t get the following, which would be useful and is what I would expect:

    Instead, in a braindead manner (IMO) forgets “urls” and takes me to just those terms:

    Now, it is possible for me to go and add that “+” in the URL, but I have to know to do it because it’s not in the UI (or a least it’s so non-obvious even I haven’t found it.) FYI, even this works:

    But it would be so much better if would make this obvious and give me links to drill down and also to include in breadcrumbs.

    Also, if they do this it would make a lot more sense for to use slashes instead of pluses, for hopefully obvious reasons:

    5.) doesn’t give any standard way to use scoped tag words without resorting to some really ugly and ultimately unworkable conventions (actually, this is a problem with user tagging in general.)

    For example, with respect to my use of “rewriting” and “advocacy” above, both really are scoped to “urls” but looks for all things tagged “advocacy” which just seems a little off to me, and even so it doesn’t give any drill downs for what things tagged with advocacy have also been tagged with:

    That said, I’m not sure how I would craft tags and URLs for scoping. Maybe they could use dashes, i.e. “url-advocacy” and “url-rewriting”, which is what I’ve tried in the past, but without built-in recognition of that from, it really becomes unworkable.

    6.) has simply stagnated.

    Most of (all?) of what offers it has offered for years. It seems like that once Yahoo purchased it they moved on to other things and let it just do it’s thing. Maybe Joshua Schachter had better things to do, but my guess was that it became part of the smaller borg that is Yahoo, and in the confusion has been left to rot by default.


    All that said, it wasn’t until Firefox incorporated as the default bookmark store that I ever really used even though I knew of it. Now I actually have the toolbar in IE7, which is the browser I use by default. I tag things constantly, but almost never go to to see my tags, or the aforementioned reasons.

    Want to see usage take off? Get Microsoft IE8 to use use for its default bookmark store (or use some Microsoft Live version of the same), and the average people’s usage will skyrocket. Address these issues in which one that IE8 supports and I’ll bet you’ll see it become mainstream.

    BTW, given where you work I’ll bet you could at least gain an audience with the IE product manager to discuss this concept, if nothing else… ‘-)

  13. (Whot, JUdell doesn’t tweet?! *grin*)

    Belated HNY and best for ’08 John; got behind the curve these past weeks.

    Laterally related, a new study [1] shows (ironically?) that folk who create formal teachings resources are themselves usually self-taught and usually engage in more informal teaching/learning processes. Which for me a) shows why I’m out of sync with them historically, since my theory is in-formed by the actualities of my experience and practice, and b) makes me wonder if they’re only (only?!) responding to market forces in what they’re producing.



  14. @engtech: I never tag things consistently.

    Don’t blame yourself, blame the tool. Nobody tags consistently (except for the hyper-excessively anal), so a “better” would make that irrelevent. Said another way, never blame the user; blame the user interaction designer for having a less than optimal user interaction design.

    Anyway, I was just using for exactly the type of reason Jon starts out his post discussing, i.e. tagging some URLs so I could send a list of items to a friend, and I identified six *more* things I think should do:

    1.) I want a thumbnail of the website so I can view a list of my URLs and hopefully remember the site where the URL came from.

    2.) I want an expandable/collapsible view for each URL where the URL’s page is displayed in an iframe below the URL. I want to be able to expand & collapse them individually as well as be able to expand & collapse all at once.

    3.) I want to be able to rate each one in the context of a tag. i.e. if I have articles tagged with “AJAX” I’d like to be able to rate 1 to 5 for the tag “AJAX”, even if I tagged the URL with something else, like “REST” (IOW, maybe I’d want to say it is a “5” for REST but only a “3” for “AJAX”)

    4.) I’d like to be able to define the order in which they are listed on a by-tag basis, and also if there are sticky ones (i.e. “always show these two first, then the rest in reverse chronological order or the rest by descending rating, etc.)

    5.) When I tag a URL I’d like to do a little comparision to other tagged URLs on the system with the same base URL and ask if I should tag what might be the canonical URL instead (i.e. if I try to tag a URL that has a marketing tracking code it would be nice for to notice that the base URL is the same and that frankly the content for the one with tracking code and the one without are essentially the same as well and have ask me which one I want similar to how Digg works on news item submission.)

    6.) And lastly I’d like as a site owner to be able to ask to re-evaluate my URLs for 301 redirects on a site, a path, and/or a invididual URL basis. For this not to cause heartburn for users, they would need to add two options when tagging: “Update URL is URL changes” vs. “Never update URL” and “Delete URL if site goes offline for 30+ days” vs. “Never delete URL.”

    Anyway, add these 6 capabilities with the prior 6 I quotes and I think’ usage would skyrocket.

  15. KISS :-) might be flawed, but it’s simple. I’m not convinced that those who ‘don’t get it’ are in that state of mind because there aren’t more bells and whistles.

  16. Occurs to me that “collaborative list curation” is solving the same thing as Twitter’s #hashtags… democratically derived categories for grouping comments/URLs. Perhaps that’s one method of describing that value prop for social bookmarking.

  17. I have to agree that the use of personal information management systems have not been well promoted, and rather, I have simply bookmarked sites for future reading, and then not been able to easily retrieve them later when needed. For one who is more interested in results than process, learning how to use the tool seems more laborious than useful. Having said this, I realize that taking time to learn the various aspects of tool, and experimenting may well be worth the time in the long run.

Leave a Reply