Joe, president of advertising revenue at Fox Networks Group, is not shy with opinions, and has emerged as one of the strongest voices for change in the advertising industry.
He's also not alone in his criticism of recent trends to sell TV ads based on marketers’ desired business outcomes, not just traditional media delivery metrics like gross rating points and cost per thousand.
Specifically, Joe asked how a TV company could sell outcomes to competing advertisers -- and also, whether it’s really possible to provide true attribution for sales of some products. In particular, he questioned the validity of whether any one spot or campaign for a car can really be tied to specific short-term sales of that car.
I’m sure you all might have opinions on what the answers are -- but here are mine:
Outcome selling does not work for all consumer categories. There’s no question that TV advertising’s ability to drive attributable short-term sales varies widely from category to category. With large-scale databases of person impression-level TV set-top-box data, it’s quite possible to have very accurate attribution for retail sales of consumer packaged goods, and be able to predictably sell those outcomes.
The same cannot be said for high-ticket, longer-term considered purchases like cars and trucks. I don’t believe that category will be very easy to attribute or sell on outcomes.
Selling outcomes to competitors means they bid up ad prices by results rather than opportunity to be seen. Selling outcomes is not how most TV companies like to operate, but that doesn’t mean you can’t sell outcome guarantees to both Walmart, Kroger and Amazon at the same time.
In the future, all TV companies must do so. It’s how Google and Facebook operate, and how they drive such great profits. They sell marketers want they want -- results -- with little risk. Those companies bid against each to get those results and keep those “acquired customers” away from each other.
That means higher prices and more inventory yield for the media owner. Adopting a performance-pricing model over time is a good thing for TV.
There’s no question that embracing outcome-based selling represents a difficult transition for TV companies. It requires different metrics. It is typically a client-direct sale. It requires different types of sales skills and sales personnel, and different accounting.
This change will be wrenching, but the end result will be good. TV companies will get much better prices than they do today, since they will be fully valued for their contributions to sales and will reshape all pricing, packaging and inventory allocations for maximum result.
This won’t happen overnight, but it will happen soon. What do you think?
If you don’t agree with how your clients are measuring your ability to help them achieve their goals, then don’t work with that client.
It's difficult to see how a TV ad seller can realistically promise a particular branding advertiser a specific sales result in exchange for a "targeted" media schedule, when the seller has no say in the development of the ad campaign---product/brand positioning or "creative" ad execution-----as well as other key factors such as distribution, the quality of the product, how it is packaged, etc. Also, it is nearly impossible to attribute sales effects exclusively to exposure in a single show or one of several networks that might be used as well as other media. There is usually very heavy overlap in audiences, hence in the effects of the total advertising effort. The only solution would be to use only one network but that would sacrifice reach.
I can see where some sort of sales guarantee might be offered for direct response campaigns, providing a mechanism for determining exactly how each person exposed to the campaign via each telecast and channel responded sales-wise. At present, attribution is simply not available for the vast majority of the audience in "linear TV" but if it was, and sales could be tracked in this manner, the most likely outcome would be for the seller to collect a large body of evidence on what response to expect for various types of campaigns---as has been done by a number of magazines---and use this as a guide in offering the guarantees. As a rule, the seller would guarantee only at the most likely result level and be willing to fall short, perhaps 10-15% of the time while hiting the goal or exceeding it the rest of the time. Moreover, sellers would, no doubt, insist on some form of compensation for exceeding the sales goal---in effect, raising the cost to many advertisers.
I seriously doubt that TV ad sellers, once they ponder the complexities and risks involved will play this game. There are too many variables---like new vs. established product, new vs. old ad campaign, competitive promotional activities, campaign wearout, etc. to warrant taking the risk of becoming the advertisers' partner, without any of the management rights and decision making authority that should be entailed.
No question J. The business is no longer about just selling media ... you've got to solve clients's problems.
CPR and that's not going to happen.
I totally agree with Dave! Solve client’s problems or somebody will. Solutions, although not all perfect, exist and we are not going back. If done right it should not cost more but less. Data drives efficiency! Free the data!
We have a plethora of attribution models. In fact it is now a plague. The majority are advertising based, and therein lies the problem. How many take into account other factors. Weather, competitors activities, price, special events. Last-touch attribution is a vanity and a relatively recent invention.
Clients calling on the various media to solve their problems and come up with solutions and (self-serving) attribution models is plain lazy marketing. The client is best placed to work out what is happening to their brand. They know their need, they have the data but not the will to do the hard yards themselves.