by Dave Zornow
Cable Avails, May 1998
Optimization is the hottest topic in national media planning and buying. If an agency doesn’t have an optimizer, they are probably talking about buying or building one.
Although a ‘90’s buzzword, optimizers aren’t new. Some of the systems agencies are installing have been used in other countries for quite some time. Optimizers are hot because Nielsen, for the first time, has made national respondent-level TV viewing data available to feed these systems. All of these systems attempt to maximize the audience that will see a schedule at least once within a fixed media budget. Some systems execute optimization on a daypart level while other more ambitious applications try to maximize a schedule’s reach by program, telecast or each individual spot.
In principal, optimizers can be a great thing for cable. Instead of using cable to flesh out a core three or four broadcast network buy, optimizers might choose cable or other media first by evaluating and choosing components based on how much incremental reach they can deliver at the least cost. Optimizers can evaluate cable by its own unique contribution to reach.
In actual practice this theory has a few problems. In an effort to differentiate themselves, each optimizer uses its own proprietary methodology. Some systems allow planners to include additional criteria, such as subjective factors like viewing attentiveness or program value (Seinfeld’s last episode might be considered a must-see TV must-have buy requiring it to be weighted more heavily than other programming.). Currently, both buyers and sellers use the same criteria – total demo GRPs or impressions delivered – to evaluate schedules. Cable sellers are concerned that they will be at a disadvantage in negotiations where each agency has different optimization criteria. So it’s true that optimization has the potential to change all of the existing rules. But with only one side knowing those rules, it’s not clear if it will it be a boon or a bust for media sellers.
Another concern about optimization is its ability to correctly predict the future. Do optimization models control for all of the factors which drive program reach? And if they do, can buyers execute the schedules dictated by the optimizers? One example of this concern is how commercial placement within a program effects incremental reach. If people tune into a program when it starts and begin to leave as they lose interest (decreasing the reach of the spots at the end of the program), the earlier ads have greater value than the later spots. But will that always be the case? What if this scenario was created by competitive programs which won’t be running the next time this show airs? What about sports events that can either build reach in the case of a close game, or suffer sharp audience declines in the case of a blowout? How about pod position? Common sense says that fewer people will watch the last spot in a commercial break than the first spot. Spot by spot optimization sounds great – but will it really work?
"We don’t yet know if an optimization strategy that works in March will still work successfully in April," says Kevin Killion, president of Stonehouse Systems and author of PCS Ala Carte, Nielsen’s Persons’ Cume Study application software, and T-View, an agency software application which uses respondent-level data to improve and evaluate media plans. "I haven't seen anything that says it is predictable. No one has proved a spot-by-spot optimized schedule is significantly better that a daypart optimized schedule."
Killion believes that optimizers are valuable media tools – but some agency researchers are taking the premise past the point of provable results. "Clearly there is an advantage to daypart optimization. But when you optimize by programs, telecasts or spots it’s much harder to come up with something predictable and buyable. I feel confident about dayparts but I’m less certain about program and telecast optimization."
Media buyers’ and planners’ enthusiasm for optimizers is being driven by client concerns over spiraling media costs. Agencies are telling clients that optimizers are a silver bullet which will improve the buying and planning process. But it remains to be seen if this really is true. "Much of the optimization enthusiasm is based on a misunderstanding," says Killion. "Many agencies are trying to do things with optimization that it wasn’t intended to do."
Finally, there is one element of the buying process which optimization has yet to address: posting. A theory which promises to improve how buys are executed but doesn’t address how buyers are evaluated is doomed to fail. To change buyer behavior, agencies and their vendors like Donovan, Core Media and MSA will need to build new posting systems that use respondent-level data instead of standard Nielsen tapes. And buyers will need to learn how to buy in a way which will both optimize and post. If EDI is any example of how long it takes to rewrite an entrenched computer application in the media business, it may be years until systems are in place to evaluate optimized buys using the same criteria planners use to create them. ##
Dave Zornow is President/TNG Research, a media research consultancy and applications development company that works with media sellers and research providers.
This article was originally published in May 1998 by Cable Avails magazine.
Return to: Home • Published Articles