Addressable television’s generally higher CPMs and scale limitations — currently about 52 million households out of about 116 million cable TV households — are no secret.
But addressable “is real, and it’s here,” and it’s in many cases producing measurable lifts in sales and return on advertising spending (ROAS), as well as in KPIs such as visits to auto showrooms, conversion to trial offers, brand awareness and ad recall, reports Daniel Solarz, supervisor, Starcom Mediavest Group/SMG.
Solarz said that SMG, which defines addressable TV as “advanced data leading to advanced delivery on the first screen,” has now conducted more than 100 addressable campaigns for more than 50 clients across its four agencies. “We’re using data from sources like Axciom, Experian and Nielsen Catalina Solutions (NCS) to identify granular targets that we can then solely deliver messaging to through [platforms] like DirecTV, Dish, Cablevision and Comcast,” he said. “We’re also linking exposure on the back end to measurement. We’re linking addressable TV to specific real-world outcomes to demonstrate the effects that it’s having on your business.”
He shared results of campaign tests for SMG clients in several different categories, and a cross-channel campaign, during the recent MediaPost Television Insider Summit.
Entertainment: Driving Viewing for Show Launches
“Addressable is a great space for tune-in and entertainment clients, because we can access set-top box data that helps identify specific audiences that watch specific types of programming,” so as to help drive viewership of new-program premieres and subsequent episodes, Solarz noted.
For example, the agency can work with partners like DirecTV and Dish to obtain set-top box data that enable identifying households that are “look-alikes” of the households that the client is trying to reach — households that are watching networks and programs believed to be similar to the new program, or within its competitive set.
“In the lead-up to the program premiere, we can serve [targeted households] specific call-to-action creative saying, ‘Here’s a new show you’re going to be interested in, and it premieres on this date at this time,’” he said. “Then we can layer in new creative saying, ‘Episode 1 was on this date, Episode 2 is on this date.’”
Clients are testing and adopting addressable because it’s turnkey, as well as capable of linking exposure to outcomes, he said. “They can look at the addressable exposure and then understand whether that same household that saw the ad later tuned into the advertised program” — which SMG agencies are doing on an experimental basis for clients.
He described the two groups for one test for an entertainment client: The “target exposed” group (households within the defined target audience that have been exposed to the ad, on a confirmed basis, through the set-top box); and the “target control” group (households also within the defined target, from which exposure to the ads has been deliberately withheld). As with any valid test, having two groups that are the same except for one variable — in this case, exposure to the advertising — enables attributing differences in results to that variable.
In this campaign, 32.4% of the audience exposed to the advertising tuned in for the premiere, versus just 4.6% of the target control households that weren’t exposed to the ad, he reported.
“That’s a difference of 600%, which is hard to believe,” he said, acknowledging that early on, seeing such results from addressable campaign tests caused SMG to question the methodology. But with similar results over many campaign tests, SMG is confident that the results are indeed “demonstrating the positive impact that ad exposure, particularly to the right people, can produce,” Solarz said.
He added that SMG also validates addressable TV outcomes by validating that it’s selected the right target audience variables from among the dozens that could potentially be used.
“One way is to compare the target exposed group, or preselected target, with non-target exposed households, or ones that don’t meet the targeting criteria but are exposed to the ad messaging,” he explained. “The hypothesis is that, after exposure, the people [whom] we believe to be in-audience for the show will tune in at a higher rate than the people [whom we believe] aren’t in-audience for that show.”
Results in this test case: 110% higher tune-in among the targeted households exposed to the ad, compared to the random sample of the population exposed to it.
Automotive: Driving Showroom Visits, Sales
Automotive is probably the largest-spending category in addressable TV, said Solarz, noting that “the value for our auto clients is that we’re identifying audiences that are in-market for their specific vehicles. We’re doing that through an array of different data sources, including Polk data and clients’ own proprietary data based on sales.”
For one automotive client’s new-model launch, the agency was tasked with addressing multiple points in the path to purchase: raising awareness, driving visits to dealerships for more information, and helping convert the visits to sales.
For the media plan, SMG worked with Acxiom to use auto-intender data, or households likely to be in the market for the particular type of luxury vehicle being launched. “We were able to identify other customers who had vehicles within that competitive set, then look at what demographics those households had in common,” so as to identify look-alike households that are strong prospects for the new vehicle, Solarz said.
Timing data were also appended on the in-market prospects, to ensure that these households were likely to buy within six months to a year — that they had not recently purchased a similar vehicle, for example.
In addition, in this test (as in some others), ads were dynamically inserted into the programming stream of the targeted households. “Ads sit on the set-top box and are only served to the household when there’s been some type of remote activity within a predetermined amount of time,” Solarz said. “So we’re making sure that an ad is being effectively viewed and therefore that [the ad] dollars are being effectively spent.”
For tracking dealership visits and linking them to addressable ad exposure, SMG engaged, through Publicis Groupe, in a piloting opportunity of technology called Mobile Audience Place Signals. If a consumer from a household that’s been exposed to an addressable TV ad goes to a dealership and is then served with a mobile ad, or has an app on in the background that provides capturing geographic location data, they can then be matched and identified as having been influenced to make that visit by the addressable TV advertising, explained Solarz.
Results: Dealership visits among targeted households that were exposed to the ad on addressable were 50% higher than those among target control households (those within the target audience that weren’t exposed to the ad).
Further, actual purchases of the new-model car were 65% higher among target households exposed to the advertising than among target households not exposed to it. That was able to be determined because the client shared dealership sales information with Acxiom, which then matched addressable exposure records to the households that bought a car.
Obviously, some households that were exposed to the ads and visited the dealership did not end up buying the new model. Many variables can influence the consumer after a dealership visit, including doing more research and looking at competitive vehicles. Or they may have decided that they’re not ready to buy a car right now, after all.
But being able to identify these households provided SMG with “some interesting insights to use for media planning,” including using the opportunity to retarget those households with more messaging in the period between the dealership visit and the actual car purchase, Solarz said.
CPGs: Driving Trial, Sales
CPGs are also among the largest addressable TV spenders at present, in SMG’s experience, Solarz reported. Reason: “We’re able to use advanced grocery retail sales data to identify households that are purchasing certain types of CPG products or are lapsed purchasers buying competitive products now, and so forth,” he said.
For example, one client’s campaign called for using an incentive to drive sampling of its product, and then converting those households to in-store transactions.
The agency worked with NCS, “which has access to a huge amount of actual grocery store transaction data and also Nielsen Homescan data, which enable them to identify households within an addressable TV universe that are in-market for a specific type of product, based on their previous purchases,” Solarz said.
Because addressable TV has finite audience reach, and NCS has transactional data for most, but not all, of the grocery-store universe, in order to expand the targetable prospect universe, NCS helped with modeling, he said. Based on their other past product purchases, NCS is able to infer that certain households are likely to be in-market for a particular product. (Households that have purchased, say, nine out of the 10 same products as households that are known to be purchasers of the product being marketed, for example.)
The agency “delivered a number of creative promotional spots to the households on the approved network list within the cable TV space,” Solarz reported, because multiple ads were needed to first make them aware of and sign up for the free product trial and then, when the trial finished, drive them to a local grocery store to buy the product.
Post-campaign, free-trial offer redemptions were matched to the addressable ad exposure. (Again, this client provided its first-party data to a third-party entity to match with addressable media.)
One result: Offer redemption was 255% higher among the targeted households exposed to the advertising than the control group.
A subsequent study of the sales was done through NCS, which was able to identify post-campaign product purchasers. A variety of sales results were able to be generated by matching addressable TV exposure data to sales data, including average dollar sales and average unit sales per household, penetration, and assessment of fair share (which competitive products gave up sales as a result of consumers buying the advertised product).
The campaign “drove an increase in the average dollars spent per household that resulted in an incremental $1.75 million in sales,” said Solarz. “This was a great example of driving consumer adoption and positive ROAS [return on advertising spending].”
Watch for next week’s Audience Buying Insider, to be published on November 25, for results from an SMG cross-channel campaign, and Solarz’ thoughts about addressable TV’s limitations outside of scale.