Commentary
Video's Changing Metrics Landscape
- by Vijay Rao , March 22, 2013
A year ago, the concept of replicating TV-like audience guarantees -- the currency of paying only for audience members, as defined by age-sex, delivered in-target -- was merely fodder for panel
moderators looking to provoke lusty debate about the merits of GRP reporting in digital.
Fast-forward today, & the jury -- major agencies and brand advertisers -- have spoken with
their wallets & ushered in the adoption of third party audience validation as an arbiter for campaign performance, with Nielsen's OCR & comScore's vCE competing for primacy.
Should we be surprised? Of course not. As agencies mandate video neutral, digital buyers are increasingly being tasked with buying online video as an extension of the TV buy. And the introduction of apples-to-apples measurement was destined to be the single biggest accelerator to knocking down the facade of buying
silos.
Theoretically, this should have represented the coming out party for online video and the potential windfall of TV budgets. But as many a publisher --
bloodied and beaten by 30% in-target delivery -- can attest, online video and the GRP have made uncomfortable bedfellows.
Recalibrating
Digital Pricing
Media 101: The smaller the audience, the more expensive to reach. Sounds intuitive? But historically (non-RTB) digital has been transacted on the
basis of a "flat" CPM, where publisher environment, rather than audience scarcity, was the overwhelming driver of price. Something that merely reinforced digital's perception of limitless
audience supply.
The fallacy of "flat" CPMs has, however, been brutally exposed by OCR and vCE. Digital buyers & publishers alike are coming to the uncomfortable realization
that the in-target CPM for Adults 18+ is very different than that for W18-34, regardless of targeting or contextual alignment. All of which leads to a fundamental recalibration of digital
pricing -- albeit, more in line with TV.
Digital's Promise of 1-1 is Overstated
The supporters of the flat CPM argue that, unlike television, digital allows you
to target individuals & therefore eliminate waste. Putting aside that you can't actually target off OCR & vCE data (reporting only), there still remains a whole ecosystem of third-party
data providers with ultra-sophisticated audience collection & targeting. Great in theory, but again cruelly exposed in the brave new world of guarantees, where the "accuracy" of these
solutions varies from 25% to 40% for niche segments. A case of the emperor has no clothes.
With Power comes Responsibility
While Nielsen
& comScore should be applauded for driving accountability, they also launched their respective offerings with scant regard for the implications to buyers & sellers, who suddenly had to
explain the novel concept of digital "waste," with little understanding for benchmarks or best practices. Even today, elements of both providers' methodologies remain opaque: What's the
direct match rate? How are unknown impressions inferred? What's the margin of error? Why is there no relationship between planning & reporting data? Perhaps most pointedly, if
you run a single campaign against both providers side-by-side, why are there massive discrepancies?
The Cost of Entry
Most agree, however, that audience
guarantees - however painful initially - have forced online video to grow up quickly and ultimately become a more transparent, standardized marketplace. Buyers are
evaluating publisher value with intelligence. Sellers are monetizing their audiences smartly & touting scarcity (much the same way a TV network does). DSPs are creating technologies to better
deliver in-target. Data providers are forced to improve audience quality and become more accountable. All these factors ulimately validate online video's ascendance into the TV mix.
So what to expect in the next 12 months? Put simply, more of the same and then some more, as Nielsen & comScore continue their arms
race. Specifically, three themes will continue to drive the discourse: (1) Both will look to supplement their panel sizes to become more representative (did someone say Google?),
(2) Mobile will become the next battleground for GRP reporting, releasing a flood of pent-up demand into video & (3) Cross-platform reporting will come to the fore,
enabling buyers to understand channel efficacy and allowing them to optimize between TV & video. Heady days.
The greatest irony of all? As online
video tries to calibrate itself around the GRP, Linear TV - via alignment of STB data & sales databases - is evolving beyond age-sex into more digital-like targeting &
reporting. But that's a separate post.
Bravo! I like this move in the industry. As a provider the biggest hole currently is "With Power comes Responsibility". It's very difficult to build effective models when the research companies will not tell you the fundamental basics of accuracy for their projection models.
Perfect synopsis of the market realities.
Thanks Vijay...This subject deserves a lot more attention than it is getting, and you guys have a great perspective.
I wonder about this though:
"Great in theory, but again cruelly exposed in the brave new world of guarantees, where the "accuracy" of these solutions varies from 25% to 40% for niche segments. A case of the emperor has no clothes."
We see the same level of discrepancy...but if I understand your characterization of it, it is that you assume that Nielson and Comscore reporting is accurate, and that the data guys provide poor targeting, (which is then exposed by OCR and VCE)...Why do you assume that? The data guys seem to have more credible data than what Nielson and Comscore use for their panels; (FB only??)...Isn't it possible that the problem is actually the opposite? And that publishers in these instances are being unfairly measured, (and paid), based on erroneous reporting from Nielson and comscore?
Luke,
Points well taken.
In my opinion it's not black or white - but rather, varying shades of grey. Nielsen & comScore's methodology certainly needs to 'mature', as does the collection practices (& accuracy) of data providers.
I guess the real point I'm making about the relative merits of audience validation, warts and all - is that it's driven accountability in the OLV ecosystem.
In the case of third party data - we've taken their claims at face value, although intuitively we've always had our suspicions about scale/accuracy.
And interestingly enough, the major data providers are now 'scrubbing' their demo segments for accuracy - at the expense of scale. Which leads to the pointed question, what were their checks and balances before?
Thanks to Vijay for a great article about the shift toward more accountable, audience-driven digital buying. Since well before our August 2010 launch, we’ve pursued open dialogue and transparency through our work with the MRC and other industry groups, as well as with our many clients and partners, including Adap.tv. We look forward to continuing those discussions to ensure Nielsen Online Campaign Ratings drives a positive impact for the industry at large and for each and every client and campaign. Happy to discuss with anyone that wants to keep the dialogue going by reaching out to me directly.
Thanks Vijay - as a client running a robust online branding campaign using Nielsen OCR as target evaluation, your article clearly articulates the challenge seen in the initial reporting. Having done this a few times now, three key items seem to have some truth for me: 1. Sites / networks have difficulty in actually serving to an age/gender audience that doesn't correlate directly to the actual site/network visitor population. (Is anyone really surprised that ESPN can reach males 25 -64 better most others?) 2. Once audience validation is added, many publishers are able to use that information to optimize and get audience delivery to acceptable levels 3. Even with maturation needed in the panels, accountability is a welcome addition. If I am going to pay more, it is extremely reasonable to expect 3rd party validation that I am reaching who I paid for.