About a year ago I started this blog with an essay on what I called 'aggregates': editorially curated bundles of content, which are what determines how successful content can be, in what light it will be seen, and which are closer to audiences than content itself (I later wrote a more succint version here) . Back then the word 'aggregation' used to mean something else, and what I saw saying was, I think, a bit unusual.
Since then the world has moved on, and similar thinking has slowly become more commonplace. A few months ago Terry Heaton wrote an influential essay on 'unbundling', in which he used the term 'smart aggregators' to refer to something like my aggregates. And at some point last year, Umair Haque published his New Economics of Media presentation, and recently he wrote (in connection to ABC's ad-supported hit-show trial) that "rebuilding is where value creation will happen...where branding will be reborn".
The latest contributions are much closer to my initial idea. In a recent post, John Hagel argues that "in addition to unbundling and rebundling of content, media companies face a choice: do they want to remain product businesses or do they want to become audience relationship businesses?" In a related note, Jeff Jarvis reflects that "the future of media is not distribution, it's aggregation" -- it having been previously established, of course, that content is not the thing either. And Charlene Li recently made a similar point.
These contributions prompt me to add some comments. But first let me insist that I am not suggesting that the people above (and some I have not quoted) have just been following my lead, or even claiming originality (of that I am not sure). I am just glad to see others pointing in the same direction.
It seems to me that in all of what has been said a key point is missing. The view that aggregates are key to content discovery is slowly becoming accepted, as is the view that a human-created aggregate is of quite a different nature than anything an algorithm can produce. But why is this so?
What is missing, in my view, is a better account of what content is, and of the role played by the context in which it is consumed--i.e. of the experience in its entirety. Media companies tend to think in terms of content because that is what they produce or trade in; but audiences, I insist, don't care about content, whether mainstream or user-generated: they care about experiences. Content consumption is part of a larger experience, which is embedded in practices that are social, and in which notions of community and belonging play a key role.
Content is nothing on its own. It only exists as part of conversations -- understood not in the usual 'blogsphere' sense of deliberation, but as shared concerns (not my term), concerns that we must partake in to be part of communities. When I buy a novel I choose it not just because I think I might enjoy it, but also because it is also being read by other people, because it is part of a larger movement that I am interested in, or because it is relevant to something else I have read. Reading is satisfactory only if I bring with me a certain baggage; and reading also adds to my baggage, allowing me to appreciate other works and, crucially, to have more of a shared background with people around me. My point is that content--or, more precisely, the transaction of consuming content--is only meaningful as part of a wider conversation that is made up of countless related transactions.
These arguments put a certain twist on the notion of 'discovery'. As many have pointed out, new media 'liberate' consumers, allowing them to 'get what they want, when they want it'. Now any publisher can reach any audience--provided audiences want the content. But how do audiences get to want what they want, where and when they want it?
Mainstream accounts of discovery posit a picture with content on one side, audiences on the other, and discovery in the middle trying to produce 'happy marriages' between the two. Crude variants speak of finding 'great' content; more sophisticated accounts add the notion of context (different people want different content at different times in different places on different devices), trust and recommendation ('if he likes it I might like it too'). But all of these accounts see content consumption as an isolated transaction, or the only important transaction; they assume that how content is reached is independent of the content itself; and that, consequently, discovery is a technical matter, important at best as a tool for securing reach. Hence all the emphasis on search, metadata and tagging--all worthwhile things, but certainly not the silver bullet they are usually taken to be.
It is easier to find flaws in the current view than it is to propose an alternative. As a first step, here a few lines of enquiry:
- The line between content and discovery is far blurrier than what the common view would suggest. In the case of video, this becomes clearer the closer we move to short-form content--which is, interestingly, just the type of material that seems to work on the web. Consider the wave of highly navigable, interactive products like Motherload, Turbonick, the various Brightcove sites, or news 'players' like CNN's or MSN's. In all of these products, the user experience is made up of a sequence of short-form bits of linear content; the user can either choose 'what he wants', or lean back as one clip is chosen for him as soon as another finishes, or alternate between the two approaches. This is how the old promise of non-linear, 'interactive' video is being fulfilled--by disaggregating long-form content into chunks and reaggregating them in interactive but editorially sensible ways. What consumers 'want' in this case is not the individual clip, but the whole navigable sequence.
- With video, much of what we call content is actually aggregates. A news bulletin is an aggregate of packages; and a package is an edited aggregate of rushes. As we start experimenting with new was of aggregating video (whether by its owners, audiences or third parties) we will see the line becoming even blurrier.
- Long-form content is different because the two hours it takes to watch a movie outweigh the few minutes it takes to 'discover' it. But even in this case we must be careful: we must break down discovery into (i) the moment I first learn about the content, which is usually offline or in linear media; (ii) the moment I decide to consume it, which may be offline or after 'stumbling upon' it online; and (iii) the moment I start consuming it, which may be straight after the decision or after a lengthy download. While with short-form content these three steps may follow each other in quick succession, this is less likely to be the case with long-form content--and here we should look in detail at how step (i) comes about. Here old-media--driven by mass-audience, non-personalised, editorial aggregates--are key.
- Editorial aggregates don't just bundle content together and recommend it; especially with short-form content they crucially set the context in which we will see the content. Consider an interactive language-learning service: a 30-second clip is meaningless without a previous context that tells you that now you will hear someone asking for directions. As another example, when a conservative blogger (a blog is a mix of content and aggregate--the blurry line again) links to a liberal one, this is not a matter of recommendation--it is a matter of setting the context in which the liberal blogger will be heard. The key word is relevance to the overall experience, not quality or compatibility. Linking is not a service (unless you are Google); it is a speech act charged with meaning.
In later posts I will try to elaborate on these points, aiming to give an account of aggregates-as-content and exploring what role they may play in the new media economy. I should admit at the outset that I am far from claiming to have all the answers; this will be a strictly tentative endeavour.