When I started this blog, this is how I began the page on About This Blog:
I believe that it is only a matter of time before enterprise software consists of only four types of application: publishing, search, fulfilment and conversation. I believe that weaknesses and corruptions in our own thinking about digital rights and intellectual property rights will have the effect of slowing down or sometimes even blocking this from happening.
Those were my “Four Pillars”. If you want read what I said about syndication early on, you can find the post here.
Since then, I’ve continued with my almost morbid fascination with observing people at work, trying to understand what work actually means in the post-industrial sense. More and more, that fascination has focused on syndication.
So what is syndication? Put simply, I think of it as the “subscribe” half of a “publish/subscribe” model. People publish things. People subscribe to things. And most importantly, the people who subscribe choose what they want to subscribe to. They choose.
Recently, when I wrote about status messages and alerts, a number of you made some very useful comments, and what they’ve done is help me crystallise some of my own thinking. [So thank you everyone who reads this blog, thank you particularly for those of you who volunteer your time in making comments so that I may learn, that others may learn.]
And this is where I’ve got to:
There’s a continuing explosion in the number of ways we create information and in the number of ways we consume information. There’re a lot of people continuing to complain about information overload, and as a result there’s some resistance to the new forms of creating and consuming information. People want filters.
Historical filters, given the constraints of the communications technologies used, were about aggregations and summaries; the aggregator or summariser played a key role in the communications process.
But there was a problem. Aggregators and summarisers, well-meaning or not, introduced their own bias, and the authenticity of the original communications was undermined.
Today we’re in a different world, with different communications paradigms, models and tools. We’re not that far away from ubiquitous, digital, persistent, archivable, searchable, communication; the costs of transmission, reproduction, storage and retrieval have all dropped alarmingly; the devices at the edge continue to improve at a similar rate.
So now we can each do the filtering.
Which made me think, how? What should the tools look like? And that took me down a whole new path. I decided to go into what makes any of us subscribe to anything at all. And the best model I could find was the Four Drivers model, developed by Nitin Nohria and Paul Lawrence (something I appear to have referred to at least seven times in this blog!).
Nohria and Lawrence suggest that there are four primary drivers to anything we do: the drive to acquire, the drive to bond, the drive to learn and the drive to defend. Further, they suggest that these drivers act in parallel and not in sequence or hierarchy; they also suggest that the power or intensity of the drivers varies continuously. I like that model. A lot.
So where does that take me? My motivation to subscribe to things published is based on my drive to learn, bond, acquire or defend. So far so good. What happens if I use that model to classify the things I subscribe to?
“Drive-to-defend” subscriptions: These are alerts I subscribe to because they tell me about something dangerous, something that threatens me or my loved ones or people/things I am responsible for. Fire alarms. Calls for help and SOSes. Urgent texts from my children. Major systems outages. That sort of thing.
“Drive-to-bond” subscriptions: These are alerts I subscribe to because they tell me something about people that I want to get to know better, people I want to get closer to. Many tweets fall into this category, as people share what they are doing. I get an idea of their interests, their beliefs, their values, and I can identify the things we have in common as well as the things that we don’t share. Relationships are often defined not by the common things but by the things unshared. And they’re enriched by both. It’s not just tweets, but many elements of the facebook mini feed fall into this category, as people join and leave groups, become friends with others, promote and join causes.
“Drive-to-learn” subscriptions: These are alerts I subscribe to because they open up new avenues for me to explore, new ways of thinking, new things to think about. Again, many of the elements of the facebook mini feed fall into this category, as people share posts, follow links, become fans, join groups.
“Drive-to-acquire” subscriptions: This is the one I’m spending most time on right now, trying to understand more about how recommendations work. What happens when someone says they’re reading something, they’re listening to something, they’re eating something, they’re watching something. They’re doing something. What happens when that’s all they say. What happens when they share more than just the event, when they share their views on this thing they’re doing. They like the book, they hate the restaurant, they’re intrigued by the film. That sort of thing.
So that’s where I’m at. The point I’m making is that this information cannot be aggregated. When Sean recommends a book to me I am interested because it is Sean who is doing the recommending; I subscribe to Paul’s tweets because it is Paul. I subscribe to Jeremy’s blog because it is Jeremy.
Not because I want to get averaged-out filtered sanitised summarised made-into-bite-sized-chunks of pap.
I’ve had pap for years. Now I want the real thing.
And I can have it. Today.
That’s what syndication is about, the power of being able to subscribe to the things I want to subscribe to. Very granularly. Out of choice and not force. With no one else filtering or summarising anything at all.
Views? Comments. More to come after I hear from you.
Since it is stated “The point I’m making is that this information cannot be aggregated.”, Do you agree that it would be useful for the author of posts to be able to state “off all the 100 posts/tweets this day/week/month, these are my top n”
Kishore, your question seems to touch on “long tail” versus “hit culture”.
I can say “these are my top n”….provided the other (100-n) are available as well. There are many ways to show the top n. Top n according to me. Top n according to comments. Top n according to visits. But in a long tail world, what is important is that we don’t revert to a hit culture and only show the top anything according to some narrow definition of top.
JP, I find the four drive model limited… If I get it right acquiring, defending, bonding and learning are “needs”. I think it’s missing of the “wants” (for lack of a better definition), for example the drive to entertain or amuse yourself. I subscribe to certain feeds or I do things the way I do at work because they make me a happier person (or developer).
Also, you write ” What happens when they share more than just the event, when they share their views on this thing they’re doing. They like the book, they hate the restaurant, they’re intrigued by the film. That sort of thing.
So that’s where I’m at. The point I’m making is that this information cannot be aggregated.”
I think the problem lies on the fact that aggregators/filters are not sufficiently powerful.
I agree that a zero to five star rank on a book at Amazon is not a true reflection of reality and looses important context information. But that’s only because Amazon hasn’t been able (or I think bothered?!?) to make a better ranking system public.
Collective intelligence applied to social data is already out there. And I am sure it’ll soon become better and more widespread, with software capable of extracting more structured information tailored to the user (using supervised algorithms, maybe?), so that you’ll still be able to follow Paul’s twitters, but – if you wish – you’ll be able to subscribe to alerts containing accurate (to you) information extracted from the twits sent during the last hour by all Pauls in Twitter, so to speak.
A late reply. The trouble with your blog is that it is quite thought-provoking and it takes me several days to gather my thoughts about some of your postings. So by the time I am ready to make a reply you have done about three more postings…
Anyway, to comment on this posting I need to comment on two things – (i)how well does the “Four Drives” model apply to syndication and (ii)your conclusions from applying that model. Now, any model is a simplification of reality to allow the user to make predictions and decisions. According to Einstein’s maxim, a model should be “as simple as possible, but no simpler”. As simple as possible, because that makes the model easy to understand and use. No simpler, because if the model is too simple it ceases to reflect reality and so becomes useless. The simplifications inherent in a model apply to the subset of reality that it is used to model.
So does the “Four drives” model apply to syndication? My gut feeling is that the model is a little too simple. It does not recognise avoidance drives (for example the drive to avoid danger (fear) and the drive to avoid boredom). It also does not recognise the drive for status. The drive for status should not be underestimated – it is significant in social networking sites and feeds, where there is kudos associated with the number of friends/links/followers you have. There is even status associated with being an early adopter of something that proved popular. (Even your blog has a “Nominate me for a CIO/IT Director award” link.)
(Later) I’ve done some further thinking on the “Four Drives” model and I’ve decided that I don’t believe there is a “Drive to acquire”. From an evolutionary perspective it’s difficult to see what advantage could have been gained by evolving such a drive. I don’t deny that there is a “Desire to acquire”, but I think this is a manifestation of other needs not being satisfied, in particular the needs for bonding and status.
But that fact that a model is too simple does not mean it cannot be used, it just means that care should be taken in interpreting conclusions from applying the model.
Now to my thoughts on using the model to classify subscriptions. My first thought is that it omits (what is too me) an important category – that is alerts to do with an immediate task at hand (that is “my plane is late” etc).
“Drive to defend”. What I called “Class A alerts” in a previous posting to this blog. Among the most useful of alerts, and normally infrequent.
“Drive to bond”. I don’t find web-based communications very useful for “bonding”. I don’t find them a very good way of getting to know people better, or getting closer to people. They have a purpose in maintaining existing bonds though.
“Drive to learn”. Well, “your mileage may vary”, but the signal to noise ratio, in my opinion, means alerts are at best an in efficient means of finding new avenues to explore.
“Drive to acquire”. I don’t understand your thinking here. For example, you seem to classify a recommendation as an acquisition. Can you further elucidate?
Martin, I see the recommendation as an aid to satisfy an acquisition instinct. Squirrel-like instinct. What makes people collect things. Have access to things.
I see the drive to acquire as explicitly not to do with “acquiring” knowledge or friendship, since they are covered by separate drivers. So I took this one to mean material things. And recommendations are useful in that process.
This post alone will make me keep reading your site. The four drivers model is not something that I had heard of before. I guess this will be a Drive-to-learn subscription. Being in the database marketing arena on a day to day basis, I have many conversations about customer segmentation and contact strategies. This post has given me more to think about than all of those conversations combined. Thanks for sharing!