Commentary
Analyzing TV 'On-Air' Promos With Online Ad Metrics
- by Dave Morgan , Featured Contributor, October 8, 2009
While many can argue about how fast we will see Internet Protocol-driven TV sets at scale, we will certainly see Internet-like measurement and campaign management impact the TV advertising market very, very soon.
Some areas of TV advertising will be impacted sooner than others. The "tune-in" business, where network and program marketers deploy "on-air" promos to build audiences for their shows, is already being impacted. In this world, viewing is "buying" and set-top-box data can already identify which viewers out of millions and millions saw promos and which of those viewers actually tuned-in to the promoted show hours, days or weeks later. The value of using this anonymous "return path" data from set-top boxes to better schedule "on-air" promos is already becoming self-evident. Here are two very straightforward examples that my scientists have uncovered recently:
advertisement
advertisement
Some TV viewers are much more responsive to "on-air" promos than others. We analyzed the "responsiveness" (did the recipients of the promos actually watch the promoted show) to "on-air" promotions by viewers that have watched just one episode of a program, but were very attentive to it (they watched at least one-half of it). When presented promos for episodes in the future, they were three times more likely to watch the show than similar viewers that did not see a promo. Yes. 3X.
Swapping over-saturated promos between sister channels can dramatically improve their effectiveness. As all us who watch television know, the same viewers see the same promos for the same shows way too many times. That is because networks use their own network inventory to promote their own shows, but only rarely promote their shows on other channels. Set-top -box data reveals that scientifically swapping "on-air" promos past points of saturation between two or more networks improves their effectiveness in actually driving viewership -- from two to 17 times. Yes. 2X to 17X.
What does this mean? It means that data is going to begin changing the way some television advertising is purchased or managed -- finally -- and "tune-in" is quite likely to be first in line. It certainly won't happen overnight, but the multiples involved are clearly too great to be ignored. 2% improvements won't move markets, but 20% or 2X improvements will. This is going to have a lot of impact in TV measurements, metrics, processes and, very likely, business models. It will certainly be disruptive to many of the incumbents -- and will also present many of them with extraordinary new opportunities -- but it will certainly be crazy getting there. What do you think?
Dave is right again, in TV as he was in the Internet, the underlying data when connected with any act of “buying” yield powerful ROI multipliers far in excess of the leverage we are used to getting out of just plain ‘opportunities to see’ data. We are seeing the same kinds of multipliers in CPG TV advertising ROI by connecting STB data to offline purchase data for the same homes.
But this has been known for years by those who have and do look at reach and frequency of promos on ones own air and/or on partner networks. What you are telling us here, Mr. Morgan, is already well known but generally not publically shared.
Jack ... I absolutely agree. To many in the industry, this information has been known for years; and some companies have built techniques and systems to exploit it quite effectively by optimizing reach and frequency exposures against specific demographic buckets of viewers, as identified by panel data extrapolated to specific shows, time slots and days. What is new I believe is that program marketers, using second-by-second census viewership data tied together at the set-top box level, can now manage reach & frequency at a much more granular level - finding highly-responsive "tribes" of target viewers in non-intuitive and counter-intuitive places, and convert them into valuable loyal viewers through optimized promo delivery. This was not possible before at this level, and most often the existence of these "tribes" were generally lost - or impossible to see - within broad demographic categorizations of panel-based data.
We should also be hungering to marry tv ad serving to wireless devices within 18 feet of the set-top box showing Proactiv ads to my teenager, or iPhone ads to me if within 18 ft. Would need an app download onto the wireless device...like facebook mobile bundled with a Comcast detectable id.
Considering most people in the biz know this information already, yet have still not actively acted on it, what makes the fact that its uncovered via a new mechanism make it actionable?
It requires more cross-network partnership, which is difficult for competitors to come to grips with, and arguably takes away audience. Unless of course its a Turner or ESPN, or NBCU, that can cross advertise its own networks. If some of that already takes place today, how effective is the cross advertising? What do the metrics show?
Well done, Dave. I agree with the premise completely, and your last statement in the comments section, to me, speaks volumes. As the type of measurements and granularity you note becomes mainstream, the need for conventional panel data is marginalized or eliminated, since the measurement will be done at least at the household level, and in some cases at the individual level. No more guesswork, no more extrapolation, no more connecting dots. Combining the data you note with true viewer TV interactivity - which your commenter Bill Harvey has been hard at work on for many years now - will redefine advertising as we've known it. Taking it a step further and combining these capabilities with the advancements in call routing for response-oriented campaigns, and you have some very exciting times ahead!
Hi Dave, your research reinforces on-air promo/break architecture knowledge that has been around for a few years. In fact, TiVo Stopwatch released a significant second-by-second live and time-shfited study of the most successful on-air promotion strategies at Promax this year.
But I will have to take issue with your point that "networks rarely promote their shows on other channels." NBCU, DCI, Turner, MTVN, ABC/Disney, and others, are all masters at cross-channel—not to avoid saturation—but to create effective reach & frequency campaigns across all their channels.
The challenge will be getting a network to run an on-air promo from another network. "Hey we're trying to get people to watch our shows why should we promote yours?"
The movie trailer world is a bit different but could apply. Everyone tries to get their movie trailer placed in front of the biggest/best-suited movies possible. The studio that made the movie gets the final 2 slots before the feature begins... Maybe there needs to be an equation that would allow for this in television. Otherwise stubborn old ways will continue.
What does all this mean for on air promos? Will we see a more "retail" approach to promos? Will unmeasurable brand or reputation style promos die out and be replaced by very specific, product / program focussed, buy now messages?