It so reminds me of Uber’s sorry not sorry ads…
I swear Tony sent me a tweet with a pointer to this piece titled Programmatic Beyond Advertising: A Not-So-Distant Future in CMF trends.
Its mainly about advertising including a bit about the just in time advertising space which is coming about because of the lightening speed of data and the ability to replace advertising/content on the fly.
Heard it all before but then there was this part…
…what if programmatic could be used for content other than advertising?
If we extend this thinking (and our imagination) a little further to consider the possible emergence of a new distribution method for cultural or editorial content based on programmatic logic and methods, we could ask whether these new “programmatic” models could be applied to the automated distribution of film and television content based on audiences and their data.
Based on this logic, “programmatic content distribution” could be imagined as a flow in which the data collected from users would trigger an automated rights transaction and content delivery process between right-holders and broadcasters. The final result would be the broadcasting of content corresponding to the preferences of the targeted user.
Yes indeed, this is the start of Perceptive Media, if you haven’t already guessed. Its always good to hear others make the same leaps in thinking of course…
Programmatic media? Don’t think that will fly as a term, I’m sorry to say. Although I have to say, this description would be more like responsive media than perceptive media.
It was in make do share warsaw that I first heard Lance Weiler talk about them in quite different contexts and it did make sense. Phil has been grouping them together as contextual media which works as a superset of both, although I worry about the previous examples of contextual media clouding things.
The next part of the article I’m less interested in but something I have thought about a tiny bit…
Moreover, it would be possible to monetize this video content by attaching superimposed or pre-roll ads to it, as commonly seen on video aggregation platforms.
This valuable collection of user data and preferences for viewing a movie or television show could be done on a voluntary basis; for example, users would simply answer a few questions on their mood, the type of movie or series, and the desired language and duration so that the platform can preselect and “program” content that meets their criteria.
But we know that the Web, which is very advanced in big data collection, is already capable of gathering this data using algorithms. Users’ actions on a given site—the keywords they search for, the links they click on, their daily search history—can indicate to the platforms what type of content they are likely to be interested in.
The problem they will get is the explicit nature of the input, I feel. Yes its easier on the web but the person is leaning forward interacting most of the time anyway. When you get into the living room it gets a little more tricky, and a implicit approach is better in my mind. Yes it can get creepy but it doesn’t break the immersion and in my mind thats very key.
The essence of the programmatic distribution mechanism would therefore be as a recommendation super-engine, more sophisticated than that currently found on various platforms.
Why is it everybody thinks fancy recommendation engines? If this is the ambition of the industry, I feel we should be breaking into another dimension. Hopefully some of the things I’m responsible for will match that ambition/moon shot.
Its intriguing to see ideas you’ve had previously, explored and implemented. I wrote a while ago about mydreamscape and how it was going to make money. One of my suggestions was product and locational placements.
Maybe a lot of people are dreaming about a certain location or a certain product. If you own that location or product you may want to own that page and make it more like yours? So for example http://www.mydreamscape.org/items/buzzlightyear/ – could be a page about buzzlight year in dreams and have images and links to the item its self. This would also be true of locations too for example http://www.mydreamscape.org/location/europe/london/thamesbarrier – would obviously link to the Thames barrier in London with information taken from Wikipedia.org and other open sources. The information architecture of exactly how this would this work needs to be sorted out.
Realizing this in my head, but decided not to include the option of having people who were spokespeople for a certain thing in there dreams. So realistically if I was to consistently have dreams about buzzlight year not only would I be featured on a item page but I’d be highly ranked. So if one of my friends was to have a dream about buzzlight year not only would they have a link to /buzzlightyear but my friends thoughts or dreams would be ranked much higher. Of course this would change once Pixar decided to own that space.
If someone checks you in to a certain place or likes a certain thing. Facebook can and will use your location/like to advertise to your friends that thing/location.
So back to mydreamscape, you would get. Ian Forrester had a dream about the Amazon Kindle automatically but the difference here is Amazon would be able to pick and choose which stories they would use in the advertising. So you don’t get that embarrassing problem, where a person has a negative dream and the advertising is based of it. Just because someone checks into Starbucks doesn’t mean they had a positive experience there, so to run it across a human eye makes sense to me.
Ok now thats out the way, I would agree that the whole process of mining users likes/checkins for data they can use for advertising purposes really sucks. But then again, to be fair to facebook. Its all in there in the EULA. If you don’t like it, for goodness sake switch to something else or stop using it.
Just quit moaning…