Commentary
Media Measurement For The NEXT New Media
- by Joe Laszlo , April 15, 2015
The vague answer is of course something to do with the Internet of things: by definition, not a single thing, but multiple things. I think the media opportunities and challenges created by smarter connected devices will be multiple, too. Yet the three areas of Internet of things where media opportunities are becoming apparent -- wearables, connected TVs, and connected cars -- share some characteristics.
Marketers are definitely interested in all three of these areas. In IAB’s most recent Marketer Perceptions of Mobile study, about 40% of marketers spending on mobile thought connected TVs would present a high opportunity for advertising in the next two years; 31% felt that way about connected cars, and 29% thought so for wearables. As these opportunities become more tangible, I expect they will share some common challenges as companies begin to measure ads and audiences for each.
Redefining Digital Measurement Standards
As we move away from the PC-based Web, fragmentation and shifting technologies will reshape how we measure media. Fragmentation takes several forms: more and more content chasing smaller and smaller audiences, but also more different kinds of devices for viewing content, over more networks, too. Media, device, and connectivity fragmentation will all require technological innovations to enable the accurate counting of audiences, and re-aggregating or defragmenting them, too.
Measurement standards from the PC era are going to need an overhaul for the multi-device world. On the PC-centric Web, counting audiences and ads relied on cookies and standards-based browsers, along with an always available, two-way data connection and a fairly unconstrained client device in terms of power consumption and data usage. All these characteristics increasingly seem unique to PCs and the world of Web 1.0 and 2.0.
To list just a few challenges:
- In connected TVs, the tendency to “pre-stitch” or insert digital ads into the content stream ahead of time, rather than synchronously, raises challenges for companies that want to count ads using IAB’s Digital Video Impression Measurement Guidelines.
- Devices that can receive content and display it, but not report upstream (which seems likely for at least some in-car services), will challenge the digital measurement status quo.
- And devices like wearables -- often just accessories to smartphones -- may not directly transmit or receive data from the Internet at all.
What’s Old May Be New Again
So what’s the new, new media measurement going to look like? Given the differences between PC digital and “Internet of things digital” it seems likely that some new connected digital media will have to be measured with a degree of old-school uncertainty.
Clearly we can’t go back to the halcyon days of TV ratings. When the next generation of measurement gurus comes into its prime they won’t believe the stories we tell them about a few thousand U.S. households decisively determining what shows were hits or misses, and how much people paid for commercial time.
That said, media measurement in the new digital era will most likely require hybrids between the deterministic approach (directly counting everyone that accessed a given piece of content, based on server records, tags, or other approaches), and the statistical or probabilistic approach (extrapolating based on a closely studied and well-understood representative panel).
PC Web assumptions that we can know exactly how many impressions were delivered, or exactly how many visits a website got, may not be possible when we’re talking about “unique wearers” or “unique drivers” or connected TV GRPs. We will have to get used to a world where advanced statistics, data fusions, and pattern-matching are vital in saying who watched (or heard, or read) what.
That’s okay. People have done business based on uncertain media data for decades. Still, this represents a significant change from the way we thought about measurement during the PC era.
Joe, while I agree that new measurement techniques should be considered and, if feasible, be developed for new media, such as mobile, your suggestion that the old TV surveys created an erroneous picture of program popularity due to their small sample sizes is not true. Yes, way back in the early 1960s, Nielsen had a national TV household panel of only a thousand homes and, on any given day, the "in-tab" number was even less---around 850. But we should remember that the average rating for a prime time show at that time was 17-18% and hit shows often reached the 35% mark. Of course, people like myself---then a very young fellow, of course----were suspicious so we checked the findings for the same shows as depicted in the local market rating reports. When these were tallied on a national basis, with aggregate samples more than a hundred times larger than Nielsen's national panel, the results were basically identical. What any new measurements should be doing for a new medium---like mobile---- is expanding our knowledge by going beyond mere audience---opportunity-to-see--- and telling us more about the effects of media and ad exposure.