Media Matters goes beyond simply reporting on current trends and hot topics to get to the heart of media, advertising and marketing issues with insightful analyses and critiques that help create a perspective on industry buzz throughout the year. It's a must-read supplement to our research annuals.
Advertisers now spend approximately $76 billion annually on television. Of this, local market “spot” buys on broadcast stations and cable account for 41%, quarterly “scatter” or periodic as-needed purchases on the broadcast TV networks, cable channels and national syndication in all dayparts make up 16%, and the remaining 43% is negotiated in so-called “upfront” deals. The upfront includes primetime entertainment fare on the broadcast networks and national cable channels, as well as early morning, daytime, news, late night and some sports sponsorships. The primetime portion of the upfront constitutes only 25% of all TV ad dollars, but it garners the most attention and is the subject of much speculation and controversy.
In recent years there has been a lot of criticism of the upfront—by which is meant the primetime upfront—driven primarily by advocates of digital media and automated (“programmatic”) time buying. Because so many upfront buys are based on seemingly archaic demos like adults aged 25-54 or men aged 18-49, critics lambaste this “outmoded” system for its lack of “granular targeting” precision. Abuse is also heaped on Nielsen for the “small” size of its peoplemeter rating panel (recently increased to 40,000 homes) and the lack of “transparency” that denies advertisers full knowledge of the sellers’ price ranges and whether their time buyers are getting the “best” deals.
While there is no question that TV’s broad-based age/sex, buying/selling demographics are very poor metrics when it comes to targeting, what the critics fail to understand is that these were decided upon as an accommodation between buyers and sellers when audience guarantees came into vogue almost 50 years ago. At that time, Nielsen was utilizing a household rating panel of only 1,000 metered homes, augmented by about twice that number of diary-keeping families. Even though average telecast TV ratings were far higher than they are now, the sellers refused to negotiate guaranteed audience pacts in cases where Nielsen’s samples might be too small to generate statistically reliable results. Moreover, since a single demo was required to determine the “currency” for each buy, advertisers with many brands pooled together in corporate purchases selected either 18-49 or 25-54, depending on the general slant of the majority of their brands. A few (less than 5%) insisted on 12-34 or 18-34 as their target, while some older-oriented buyers opted for 35+; but approximately 90% of the buys were made on 18-49 or 25-54 metrics. This is still the case.
Critics of the upfront are stymied by the fact that so many national upfront negotiations represent the combined requirements of all of an advertiser’s brands, often including 20-30 brands in a variety of quite different product categories. Anyone can see that even if each brand’s targeting definitions were taken into account, the only way buyers could factor these preferences into a single corporate buying currency would be to average them. Unless most or all of the brands had virtually the same consumer user signature, the resulting average demo would probably be no better than 18-49 or 25-54.
Another problem that critics of the upfront must confront is their failure to appreciate the true relationship between buyers and sellers. Most national advertisers are wedded to TV as their primary communications platform. While digital media and other options—magazines, radio, etc.—are available, many advertisers believe that they must be strongly represented on national TV, despite its rating fragmentation and commercial zapping issues. And the sellers know it. Except for an occasional “buyer’s market,” caused by depressed economic conditions, the sellers rule. They may make concessions here and there and occasionally grant favors to friendly customers; but overall, the sellers dictate the way time is sold. And they have no intention of turning over their masses of GRP inventory to eBay-style open auctions where everyone knows what the prices are, and the buyers can cherry pick time in whatever programs they want, whenever they decide to make a purchase.
Because they manage a perishable inventory of GRPs and must garner the maximum ad revenue yield across all of their shows, TV sellers offer time in discounted package deals, mixing in the most preferred programs with their many so-so entries and outright flops. This succeeds because the cost per viewer—even a meaningfully targeted one—is lower than what the buyer would have paid had only the “best” programs been available. The sellers make certain of that.
From an advertiser’s standpoint, the upfront offers five important benefits:
Access When You Need It
Upfront buying is done in the context of the anticipated media planning for all of a corporation’s brands in the upcoming season. Some brands have a preference for certain program genres and wish to ensure that time will be available to them exactly when it is needed in such contexts. In the quarterly scatter market, you take pot luck in CPM pricing, in the types of programs where time is available, and which telecast dates have openings. Buying in the upfront guarantees the corporation timely access, at a known price, in the kinds of program environments its brands favor; scatter does not.
Many national advertisers are very concerned about the “look” of their TV buys, especially on the broadcast networks and some of the larger cable channels. Such advertisers are often image-conscious, and wish to be seen as sponsors of prestige enhancing TV shows. With this is mind, they feel the need to promote their “big show” sponsorships to the Wall Street analysts and, most importantly, to the trade: the store chains, dealers, franchisees, etc. that actually sell their products to consumers, plus the trade press that services each segment of the distribution chain. You can’t proudly announce that you are a “sponsor” of The Big Bang Theory or Empire when time in such programs may not be available to you in the scatter market.
Using Corporate Clout
Many large corporations recognize that allowing each brand to go its separate way would play into the seller’s hands regarding both pricing and the types of shows they would advertise in. They also like the idea of having huge agency media buying shops orchestrating massive upfront deals at generally favorable CPMs. Even if these are not specifically targeted to suit the needs of each of their products, these can be satisfied to some extent later when the buys are allocated to the various brands. Sometimes this is accomplished by computerized models that attempt to optimize the demographic fit of each brand’s GRPs; often the CPMs are manipulated so brands with a poor fit pay less per viewer impression that those with a good fit, thereby equalizing the “value” of all of the corporate buy’s exposures across all of the brands.
Playing the Market
Playing the market is also an important part of the buying/selling game. Many corporate advertisers, along with the sellers, gamble by withholding smaller or larger amounts from the upfront each season, in an effort to capitalize on anticipated CPM differentials in the scatter market. If a corporation allocates 75% of its national TV budget to the upfront at a $10.00 CPM and spends 25% in the scatter market at an $8.00 CPM, its net gain for the entire budget is a 5% lower CPM compared to an all-upfront purchase. This savings can augment each brand’s audience delivery via added scatter buys, or the money may be reallocated to other uses. Hence, a stable and predictable upfront buy creates a strategic framework for possible short term tactical gains by going the scatter route. Even if the scatter market suddenly goes sour, the bulk of the advertiser’s TV budget is secure.
Finally, advertiser media mavens and agency buyers enjoy the upfront process, with its pilot screenings and gala presentations, and they get a feeling of participation in criticizing the proposed program schedules, strategizing about how to approach the sellers, etc. In short, the upfront is fun, while scatter is a mostly by-the- numbers grind.
The sellers’ primary benefits from the upfront are two-fold:
Minimizing Programming Risks
The broadcast TV networks and, to a lesser extent, a number of major cable channels, spend upwards of 60% of their projected ad revenue incomes on program development and procurement, much of it through outside suppliers. They need the security that massive upfront ad sales provide to front those portions of their program costs that must be spent in advance of the season, as well as payments they must make to the producers as the episodes are delivered.
Controlling The TV Marketplace
Unlike the advertisers, each of whom operates separately, the major sellers see much more of the activity and they can gauge and manipulate the availability of GRPs and their pricing accordingly. This is particularly true for the broadcast networks, which usually make the first major deals, thereby setting the CPM norms for most ensuing buys. Another advantage of the upfront is that it ensures a seller the future benefits of past rating success. When a network has a good second season and goes into the upfront with rating momentum, buyers are, quite naturally, influenced by what they have seen in the most recent Nielsens. As a result, the network may lock in high CPMs for the next season in anticipation of maintaining its higher ratings. If the ratings later falter and it is obliged to provide the buyers “make good” spots to ensure audience delivery, the network still gets paid more per viewer “impression” for a full season.
As we have noted, there is much clamor and hand-wringing about “reforming” the upfront or, at the very least, making significant improvements to the process. Of the two, the “reformers” have had the least success. Invariably, they propose something akin to open-bidding systems—à la eBay— where everyone knows exactly what’s been offered and then bids. Not surprisingly, the sellers have not been supportive of such an idea, not only because many offers are made as package deals, but also because the sellers do not like the idea of “transparency.” How would this help them?
Still, some progress has been made, chiefly on the metric front. About ten years ago, the agencies were successful in getting the networks to agree to use average commercial minute ratings, not all-content ratings, as their buying and selling currency. The networks countered by insisting that delayed viewing (live plus three days, or “C3”) be the basis. And in an effort to add a few percentage points to their audiences, many buys are now expanding their time frame to include 7-day delayed viewing (C7), and there is movement towards including exposure on digital and out-of-home venues, which will add still more viewers, albeit often on an extended delayed basis for the digital component.
Under pressure from clients, the agencies have also developed add-on indices that are combined with Nielsen ratings and reflect the degree to which various shows “engage” their audiences. Since engaged viewers are more likely to watch commercials, this innovation is presumed to have improved the average advertiser’s commercial exposure rate. Unfortunately, the sellers’ various discounting and bundling schemes generally defeat the engaged viewer concept. Buyers are forced to accept highly-engaging and not-so-engaging shows sold together to get the lowest CPMs, and that’s exactly what most do. The proof is in the pudding; TV commercial recall levels have not improved in recent years, despite the assumption that advertisers have found a way to improve their commercial exposure levels.
Recently the networks and selected advertisers have begun to use “third party” product buying information interfaced with “big data” set-top box TV set usage ratings to create supposedly improved targeting indicators. Thus, it may develop that homes viewing TV Show A are 20% more likely to buy a particular product than those tuned in to TV Show B, and such indices would then be applied to Nielsen’s viewer ratings as a way to improve the time buyers’ targeting capability. It sounds fine in theory, but the “big data” set usage ratings that are used to create the indices do not indicate who is viewing, and people—not TV sets—are the advertisers’ real targets. TV set usage tallies invariably create the illusion that younger, more affluent homes, which have more screens and residents and therefore higher set usage, watch most shows to a greater extent than older folks and lowbrows. In reality, exactly the reverse is found in studies of viewing behavior. As a result, it’s highly suspect to employ set-top box data to accurately define the product use patterns of TV audiences. In addition, as with the engagement metrics, the seller’s ability to bundle many shows together in discounted packages tends to nullify whatever targeting efficiencies might be gained by cherry picking selected programs, but at higher CPMs.
Is there a better way? Despite all of the failed attempts to improve TV time buying by tinkering with the ratings or by adding new metrics such as engagement, it is possible for advertisers to get more value from their upfront purchases while operating within the current system. The basic idea is to strive for an opportunity to attain a better brand-by-brand fit, rather than accepting whatever demographic imbalances—mainly too many old viewer impressions and too few younger ones—that are inherent in most upfront buys.
Here’s how it might work:
Defining Brand Targets
As before, each brand would stipulate its GRP requirements by network type and daypart for the corporate buying “umbrella” demographic, say adults 18-49. In addition, each brand would provide an index breakdown by demographics, indicating how important each one is in terms of potential product sales. For example, Brand A might require 5,000 adult 18-49 GRPs on the broadcast networks’ primetime schedule but its indices by age and income could look like this:
Applying these indices to the overall 18-49 corporate buying GRP goal (5,000 GRPs), the advertiser’s media director would set a GRP target for each demographic cell. Hence the 18-34/ upper income GRP goal would be 6,000 (5,000x120), the 55+/low income GRP goal would be 2,000 (5,000x40), etc.
When all of the corporation’s brands were assessed in the same way and the findings totaled, the result would be an overall adult 18-49 GRP goal for the buyers to attain at the lowest possible cost. In addition, the buyers would be charged with the goal of coming as close as possible to the desired corporate GRPs for the most important demographic cells, in this case the younger, upscale ones.
For example, let’s say the buyer is considering three proposals, each for 1,000 18-49 GRPs in a given quarter. All three proposals come close to the requested GRP amount, but as shown in the accompanying table, Proposals A and B tend to peak their GRP delivery among older lowbrows to a considerably greater extent than Proposal C, which is still skewed in that direction, but provides a relatively more balanced GRP performance than A or B. Yet A’s and B’s CPMs against 18-49s are seemingly superior (lower) than C’s (see Table).
But what happens if the buyer applies the marketing value weights described above against the GRP delivery by demo cell, in effect calculating the marketing value of each GRP by where it falls by age/income groups? The resulting calculations, as shown at the bottom of the table, have Proposals A and B in a virtual tie, each generating about 823-830 “effective” GRPs, while Proposal C, despite a somewhat higher 18-49 CPM, supplies 885 effective GRPs, a 7% improvement over A and B.
The same approach could be applied as the buyers evaluate the incremental value of each proposed buy against the total weight of those packages to which they have already committed. Given that the goal is 6,000 18-49 GRPs, and 4,500 have already been bought (taking into account their weight by the various demo cells), what does another layering of buys add in terms of attaining the corporation’s demographic goals by age/income, and at what cost?
Processing The Data
Needless to say, the time buyers and their media research cohorts would have to prepare average minute rating estimates not only for adults 18-49, but for each of the demographic cells for all of the contending sellers. This is certainly feasible, using past Nielsen findings as prototypes for the future, including new series that have not yet been measured. A computerized model would allow each package from each seller to be analyzed for its total 18-49 GRPs and their demographic breakdowns relative to cost. As each buy was finalized, the advertiser and its time buyers could see how they were doing in terms of GRP goal attainment relative to the dollar amounts committed.
Making The Buys
Clearly, it is unreasonable to expect the buyers to exactly match the corporate GRP goals across an array of demographics, no matter how hard they try. Also, the overall cost efficiency penalties imposed by going overboard in attempting to attain the requested GRPs in certain cells may be too great to bear. However, unlike the current practice whereby the buyers and sellers know exactly how their offerings are being evaluated, this new system would, for the first time, give the buyers a hidden edge. The sellers would not necessarily be privy to the detailed cell-by-cell breakdowns. As the buyers ask for more and more revisions in their proposals, the sellers’ packaging experts might give more than they realized, trying to win the business.
Ultimately the advertiser, in consultation with its agency-of-record for a TV buy, must decide when to call a halt to further negotiating and whether too great a price is being paid to cater to difficult-to-attain demographic audience compositions. Also, the buyers may regard the whole idea as impractical and express concerns about their freedom of action, particularly getting quick approvals to finalize buys where the packages and CPMs look reasonable. Indeed, it may behoove advertisers who opt for a better demographic fit to lay in core buys at maximum CPM efficiency—say for 50-60% of their budget—under the old method as a failsafe mechanism. The remainder of the budget might then be tinkered with to see to what extent the demographic imbalances can be whittled down and at what cost.
Is this a magical panacea that ensures a better brand-by-brand fit for every upfront advertiser? Of course not, but it does give the advertiser and its buying arm a much better chance to at least attempt to satisfy the brands’ collective targeting needs. Sometimes it may pay real dividends, sometimes not. In a recent consulting operation, we evaluated an upfront buy done the usual way—using a single conglomerate demographic target—against what it might have been if an attempt had been made to tailor the GRP delivery to more closely match the all-brand consumer profile. In this case, there was a loss of about 5% in overall efficiency, but the average improvement among key demos of importance to the brands was 9%. Isn’t the possibility of such a trade-off worth exploring?
In closing, we urge readers to consider the alternative to TV’s upfront, in which several major corporate buys account for most of the upfront action. Without it, we will have each brand individually trying to negotiate whenever the need is felt; that’s upwards of 7,000 brands going it alone, many with different targeting metrics and often buying one month or quarter at a time. How could the sellers handle so many negotiations and subsequent renegotiations? How could the agencies make so many different buys and service them? The answer would be by greatly expanding both the sales and buying staffs and charging advertisers far more than the current buying fees for national TV. And how does the seller maximize total ad yield and profits when so many varying metrics are employed and so many players are involved? It would become nearly impossible for them to manage their inventory. In our opinion, the result would be chaos.