January 25, 2023
GroupM Lowers Ad Outlook, Scanzoni Calls Nielsen Panel Outmoded
- by David Goetzl @dgoetzl, December 1, 2011
GroupM has downgraded its 2011 and 2012 projections for ad growth in the U.S. Chief Investment Officer Rino Scanzoni said the previous projection of a 3.8% growth in 2011
for all measured media now stands at 3.3%.
Looking ahead, the 2012 forecast had called for 4.2% growth, which is down to 4%. The altered projections were released at mid-year. The predictions use GDP growth as one basis.
Scanzoni spoke at an industry event Tuesday where he delved into two areas involving TV measurement he had referenced before.
Like many, he suggested Nielsen's national TV panel is outmoded and questioned its long-term viability. “A sample-based system doesn’t apply in the world we're in,” he said.
It may have worked when there were three networks, he said, but not 80-plus cable channels and viewing on other devices. Scanzoni would like to use some melding of set-top-box data with Nielsen's panel. "I don't think it's going to be one versus another," he said.
advertisement
advertisement
Nielsen is pursuing a so-called hybrid model, but not in the national marketplace, as Scanzoni would like. GroupM's parent company WPP owns Kantar, which has a measurement system based on set-top-box data from DirecTV.
Scanzoni also pushed for the local TV market to move away from "live only"
ratings as the currency, since it fails to take into account time-shifted viewing and DVR-enabled ad skipping. He said it is "not good business practice," when an estimated 50% of ads are
skipped.
Separately, Scanzoni is eager for Nielsen to begin measuring viewing on iPads, which it says is in development. “The question is trying to get them to do it quickly
… it takes them a long time to do things that need to be done and that holds back the industry,” he said.
3 comments about "GroupM Lowers Ad Outlook, Scanzoni Calls Nielsen Panel Outmoded".
Read more comments >
Next story loading 
I find it incredulous that Nielsen is still using the old "tried and true" sample based reasoning.
Nielsen was old 10 years ago and as marketers we still hang on to dysfunctional methods.
We have so many bright young minds in this business that are unfortunately harnessed by the dollar cost of change. No one wants to pay for it....not the agencies nor the marketers. But some smart young group in a garage may soon figure it out and out-do Nielsen!
Nielsen is far more interested in holding on to its antiquated but entrenched business model with as little investment as possible than actually delivering an effective version of the service it allegedly provides.
Truly, this is an "Emperor has no clothes" situation. Few want to call out the central absurdity of a few thousand households still dictating what's a success and what's not, because it will reveal the profits of several different industries are built on sand.
The more media influencers and leaders, like Mr. Scanzoni, challenge the status quo, the better off we'll all be. In the long run, anyway.
Clearly a 'sample-only' approach is reaching the end of its lifestage. But this actually has little to do with this being 'sample-only' research, but is more a reflection of reliance on single-data sources only (of which sample-only is one model) in an increasing fragemented media landscape.
One only has to look at the initial reliance on the 'nirvana' of 'census-based tagging' in the online world, until it began to dawn on users that this approach dramatically overstates audience (how convenient). The approach being adopted in online is to use tags to measure the traffic (i.e. potential volume) and then to meld that with sample-based behavioural data to turn that traffic data into believable audience data.
The same model works for TV with RPD providing the traffic, and the panel providing the audience data. Reliance on RPD as the sole source basically moves the measurement back to the days of household ratings - hardly a quantum leap forward in more accurately measuring a metric that is old hat. But by blending this with the panel data we begin to be able to quantify the tough stuff - e.g. duplication of viewing across multiple devices in the home, audience profiling, number of viewers per tuned device, audience accumulation over time.
Replacing panel data with RPD-only would be as poor a decision as not utilising RPD data with the TV panel.