| ||||||||||||
PRINT
COMMENT
SUBSCRIBE
RSSIt's a call that most of us (myself included) would heartily endorse as the prospect of wholesale and routine access to such data continues to move closer to reality, yet still remains tantalizingly out of reach (with a few notable exceptions).
Obviously we are getting there, but as John Grono of GAP Research observed in a response to Tracy's piece, even when we have all the set-top data we could wish for, we still won't know whether or not content is being viewed (unless channels are constantly being changed or a return path is active and enabling us to log different interactions with the content) or by whom.
No set-top box is going to be staring into the home in "Truman Show" fashion, recording every second of behavior in front of and near the TV to tell us when and for how long people are out of the room, when they are using other media, talking, sleeping and so on. This is an issue we addressed at Ball State in a pilot research project called Remotely Interested and in a white paper called Engaging The Ad Supported Media.
Leaving aside the cost and privacy issues of enabling electronic surveillance through the set-top, these days you'd have to wire up the entire home to account for the multiple TV sets typically found in the home.
But the other problem with set-top data is that -- valuable as it is -- it ultimately only tells us about TV, thereby helping to perpetuate the silo-based approach to media research and measurement. Although TV remains the primary medium in terms of time spent in the average day, still the ways in which we can use the TV and the number of video-capable platforms continue to proliferate -- so even set-top data will leave questions unanswered. And we'll still have the perennial problem of trying to meld together disparate sources of data intended to serve different media, as we search for a clearer picture of media use in context.
Often it's like trying to wade neck-deep through a data swamp that keeps pulling you down with no obvious way out.
As content becomes available on more platforms and as content owners seek to monetize it accordingly as the audience delivered by each platform grows, so the need to understand the way in which each platform is interwoven within people's daily lives increases. Yes, we still want better and more accurate measures of media use by platform, but unless our understanding of how they relate to each other improves also, we will be increasingly out of step with the silo-free consumers we seek to understand and reach. After all, when was the last time you encountered a single-silo consumer in this cross-platform, multitasking world of ours?
Of course all this is easier said than done. Adventures in cross-media research are still in their relatively early days (aside from one-off, generally proprietary projects). Where initiatives are underway, they typically revolve around just a couple of media rather than a more comprehensive mix of the media available to consumers. Some rely on self-report -- which, when you start asking questions about the use of several media, become very shaky indeed (to understand just how shaky, check out the first Middletown Media Studies report, which examined the difference between phone, diary and observational findings).
Naturally, there will always be a meaningful place for research that addresses specific issues relating to the consumption of individual media, but for those of us seeking a broader and deeper understanding of total media use (and it's context), there are few options available.
This week's announcement by Nielsen and the Council for Research Excellence of the decision to fund the Video Consumer Mapping Study will result in a very significant contribution to an understanding of the consumption of video across all platforms (full disclosure -- the study is to be implemented by Ball State and Sequent Partners). But while the observational method it employs will capture video use across all platforms and locations as well as the behavioral context in which it occurs, the method is necessarily costly and is ultimately most valuable when triangulated with outputs from other research -- something that will undoubtedly be done with the study.
This suggests that data fusion will have a role to play in our continued pursuit of media enlightenment. And yet -- unlike in Europe -- fusion seems to get a mixed reception here in the U.S., so the rate at which it becomes an accepted part of the process is uncertain.
One thing is certain, however. As complex as the media ecosystem is right now, it is only going to get more complex - and at an ever-increasing pace. What looks challenging and difficult to navigate now will probably look like the land of milk and honey when we look back in ten years' time. As such, it is equally certain that no one solution is going to provide us with what we need to optimize communications efficiency. We will increasingly go digital in our measurement systems, but we will still need to understand the context of media use that data harvested from devices and meters will never capture.
As goes the complexity of media, so goes the complexity of research and measurement. And unless we truly master the latter, we'll never be able to make best use of the former. Instead we'll find ourselves ever-deeper in the data swamp - loads of information but no real way to assess reliability or extract value from it.
1 person recommends this article.
RSS FEEDS
Media enlightenment is not so far from reality as you indicate. TRA’s method is inherently cross-media and our customers will be running three screen reach/frequency and ROI reports soon. We’re excited to be working alongside you and Tracey in parallel paths with an eye to putting it all together in the months and years ahead.
Methods allowing aggregation across platforms and (eventually) measuring the direct behavioral impact of viewing of a single campaign across multiple platforms could erase a huge portion, if not all, of the billions of dollars left on the table annually because inventory that can't be measured accurately can't be sold at full value.
You should look at what Backchannelmedia is doing. They're getting right at the cross-platform conundrum -- not precisely in the way Mike is contemplating - but with an approach is both elegant and plausible (and, frankly, cool).
Incidentally, I don't see 100% accuracy of measurement as the holy grail. Measurement is just a component we use to value one campaign and make the next one more efficient. We want to get more value out of the inventory. It's better to know if a set is switched on or switched off, but 100% accuracy of that measure won't tell much. Better yet to know that it's on and what channel it's tuned to; better still to know if anyone is watching, and of course we'd prefer to know who that is. But what we really want to know is the extent to which a particular impression on a particular platform (or a multi-platform campaign) directly influences behavior and attitudes (which are proxies for future behavior -- the "long tail" of advertising). There is no 100% accuracy. It's an infinitely iterative process of better measurement, better insight into what to measure, and how to apply those evolving data sets to be more effective with future inventory.
Using a household sample of one, based on channels/programs viewed, a jr. level researcher would be able to ascertain rather quickly, which of my two teenage children, my wife, or myself, was likely to be viewing one of the two sets hooked up to HD-DVR boxes in my abode.
100% accuracy, certainly not, but how much accuracy is our industry willing to fund?