Here Come the TV Outcome Wars
Can any one player really win?
Forget the tired Currency Wars - here come the TV Outcome Wars.
Even as NBCUniversal is going at Nielsen in a very public fashion, and multiple media giants grumble about the delayed Gauge reports, the researcher’s credibility hits don’t seem to be shaking up its business much. Nielsen keeps signing new deals, and will serve as a key data provider for the upcoming upfront season.
Meanwhile, most everyone in the TV business -like the advertising world at large - is talking up, or under pressure - to deliver outcomes.
This in a medium where that has never been easy. Which may be why we’re seeing so much activity of late, such as:
Viant acquiring the TV ad research firm TVision
The outcomes-focused analytics company EDO rolling out a new AI product in the market for free in an effort to gain agency traction
The would-be Nielsen challenger iSpot launching an ‘Outcomes at Scale’ product a few weeks back
Not to mention Pinterest acquiring the performance TV tech firm TVScientific late last year
And of course, Nielsen introducing its own Outcomes Marketplace last summer through a partnership with the firm RealEyes, which tracks everything from facial expressions to emotional reactions to ads
All these entrants seem to be aiming at making TV work a bit more like Google or Meta when it comes to trackability - which isn’t easy. “TV is tricky from an attribution and measurement perspective,” said Jason Fairchild, CEO of TV Scientific. “You can’t click on a TV. Last click is sort of out the window.”
But does everyone understand how TV outcomes should or shouldn’t work - or where they are headed? Here are some huge overarching questions (in my view):
How exactly should you measure TV-driven outcomes? Over what windows of time?
Can outcomes even be standardized (like a Nielsen rating, for instance)?
Should they fundamentally be tracked by third parties, a la Nielsen or comScore? Or will they always be individual to brands and media companies?
Should they be tracked after the fact - or baked into ad transactions?
On the standardization question, Moe Ismail, Senior Vice President, Product Management, Epsilon said that uniformity is unlikely. “Currency works because it answers one question,” he said. “Outcomes answer many. And those questions vary by advertiser and objective.”
Brands are optimizing for different behaviors, from store visits to sales to installs, which makes a single outcome provider unlikely. The advantage will go to companies that can consistently recognize individuals, unify data, and connect exposure to real outcomes.
On the neutrality question, Viant CEO Tim Vanderhook said “most platforms are currently measuring themselves, which creates inherent bias.”
“By bringing TVision into Viant, we’re not compromising neutrality, we’re scaling it. TVision’s data remains market-wide, measuring attention and co-viewing across linear, CTV and walled gardens, so advertisers can finally understand what’s actually working across the entire ecosystem.
“A traditional data partnership would have kept attention where it’s always been, after the fact and disconnected from decisioning. The real opportunity here is bringing that data directly into the bidstream so it can inform planning, buying and optimization in real time.”
Of course, Viant - while ascendant, is only used by so many brands. The Trade Desk is still a major DSP player, and increasingly, many are turning to Amazon’s DSP to buy TV inventory and measure its performance. Few companies can match Amazon’s identity and transaction data - and given that Amazon now has deals with Netflix, Disney and others makes it a formidable player in this attribution race.
Of course, not every product is sold by Amazon - or lends itself to this kind of immediacy. Another contender would mobile attribution companies like AppsFlyer or Kochava, which come at this business from a very different angle. Kochava CEO Charles Manning talked about this in a recent episode of Next in Media.
CTV is “the app economy all over again,” Manning said. Kochava is plugged into the operating systems of many big TV makers, including Samsung, LG, Vizio, etc.
“Outcomes absolutely matter,” he said. “The view-and-do combo is when you have an immersive experience on the TV where you’re viewing, and then where you’re doing is on your mobile device - a measurement solution has to account for that dynamism.”
That connection would seem to be where EDO wants to play - correlating searches that happen soon after TV ads air, for example. While EDO’s new AI tool is available for free - its core outcomes product is not.
“At the bottom of the hierarchy of values sits the mere observations of ad intelligence. It’s the data exhaust flowing out of the TV industry—straightforward and necessary, yet costly and time-consuming to gather,” wrote EDO founder Kevin Krim of the free offering in an Adweek editorial. “Further up the measurement pyramid are proprietary audience and outcome measurement data that ultimately power optimization at the pinnacle. “
For its part, rival iSpot is so fixated on outcomes (versus say, trying to beat Nielsen at currency) that it hosted its own closed-door summit on the topic this week.
The company - which has sued EDO, emphasized that tracking outcomes in TV takes a while. Just tracking what someone does right after seeing an ad - even with phone in hand, is just one signal.
Per iSpot:
Fewer than 50% of conversions occur after only 1 exposure,
For most industries it’s less than 25%.
It takes an average of 7 to 10 exposures before a consumer takes an action
That range of data speaks to the unique nature of TV. Not only is it primarily (but not always) a lean back experience - the response to an insurance ad is going to vary wildly compared to a less considered, or more urgent purchase.
CTV is “absolutely performative,” noted Manning. “It’s a function of the how and the approach.”
So is the approach to look at immediate searches following TV ads? To only buy ads that deliver certain levels of attention, or emotional responses? To measure everything people do on their phones while watching TV? To measure instant responses, or over 14-day periods?
Right now, the answer seems to be yes to all. “TV attribution is complex because exposure and outcomes live in different places,” said Epsilon’s Ismail. “Connecting them requires consistent recognition of the same individual across environments. That does not mean it stays fragmented. Advances in identity and data collaboration are making it possible to unify those signals into a more complete view.”
Sounds hard. But war is hell.