Four channels. Four measurement suppliers. Virtually all media consumption and advertising covered. What a simple world.
No more. Digital disruption, the rise of performance advertising -- and, most importantly, massive and accelerating audience fragmentation across an expanding number of media channels, suppliers and devices -- are to blame.
Media measurement has been in the news a lot lately -- largely centered around the granddaddy channel, television.
The TV networks and their trade group the Video Advertising Bureau are publicly railing against Nielsen for its recent, quite public, measurement failures.
Nielsen has responded by trying to put its Media Rating Council accreditation on hiatus and focus its energies on building out a more holistic measurement platform for the future, Nielsen One. And no one involved seems happy or likely to be content any time soon.
Lots of folks have been asking me for my thoughts on what this all might mean for the future of the media measurement market. So here they are:
Nielsen’s TV measurement is not going away. No question that Nielsen has some real challenges with its TV panel, but there's no simple way to replace it as the currency for the majority of national TV ad spend in the U.S. As research legend Jack Wakshlag taught me years ago, being an accepted media currency is about much more than just having accurate numbers. It is also about comparability, stability and trust, and it typically takes decades to establish those capabilities across a marketplace with hundreds of sellers, thousands of buyers and 10+ trillions of ad impressions.
For better or for worse, the C3, C7, gross rating point and sex-age demographics are etched in stone in the contracts covering tens of billions of spend for those trillions of impressions. Nielsen is the only company right now that has a panel, as inadequate as it is, that covers all of the types of TV viewing -- broadcast, cable, satellite and streaming -- and is balanced for all demographics, particularly minorities and those in homes without fixed broadband (more than 20% of U.S. households).
Many emerging measurement companies are going to pick up a lot of new business. There are many digital-born players with media-measurement products in line with where media consumption and ad spend are going. iSpot.tv has been a leader in TV and video ad campaign and outcome measurement for years, with massive reach across U.S. households. TVSquared has been a leader in conversion tracking. Samba TV and VideoAmp are bringing even more real-time TV measurement to the market.
Truthset is upping the game on data validation and accuracy. TVision Insights is teaching us a lot about who is actually watching TV and how engaged they are. 605 measures the impact of TV ad campaigns with return-path data. LiveRamp's Data+Math powers and measures campaign-level targeting and outcomes on TV. And there are many, many more.
Net. net. The future of media measurement will not be like the past, with a few simple numbers from a few monopoly suppliers of currency data. Nope. The market will be much bigger, with massive, growing demand for tactical measurement platforms that can operate in real time and help marketers, agencies and media owners better use and measure their many different media channels and the nuances of each, leveraging the kinds of targeting, optimization and real-time measurement and reporting that are table stakes in all the major walled-garden digital platforms.
So -- Nielsen isn’t going away any time soon, but that doesn’t mean that there aren’t going to be a lot of new market opportunity for a lot of new media measurement suppliers.
For sure, not only will the media-measurement marketplace grow, but it’s going to fragment -- just like all of the media channels and the audiences that watch them today.
Not only is and was the Nielsen "panel" innadequate, as you state Dave, but during Covid, it has been reported that they stopped sending their field force representatives into their panel homes.
Over my years of service with MTV, NICK, Comedy, Univision and many other companies, I frequesntly asked Nielsen a lot of questions. However, because I am not currently a Nielsen paying customer, I cannot learn when they stoped or for how long this occured. Why would a field force employee meed to go into a home? Many reasons. To hook up any device that gets "TV" that the home buys or adds. To fix any "defaults" -- where a device was not sending appropriate viewership data. To disconnect a home, whose time in the panel ran out, or simply because the HH wanted to leave the panel. Or, to add new homes to the panel, given that many homes left the panel.
Nielsen, rightly, takes many HHs "out of tab" (and they do not tabulate any viewer data from these "out of tab HHs") and the rumors are that the number of out of tab homes increased and grew large. How large? Again, I am not a customer, so I dont know. I do, however, wonder what percentage of panel HHs are "in tab" right now.
Typically, and logically, HHs who are "out of tab" are the highest viewing HHs with many more devices (that can faiult) and manu more HH members than the average HH. This alone could have had a devastating impact.
I agree with most of what you are saying, Dave, but there was always competition in audience measurement---alternate sources and, often, rival companies splitting the market for exactly the same kinds of data. For example, in the local market field for TV, The American Research Bureau---later known as Arbitron, was the leader throughout the 1950s but Nielsen then fielded its own rival service and the two duked it out for many years before Nielsen gained the upper hand. Likewise in radio Hooper and Pulse were the local market services which were supplanted by Arbitron in the 1960s.There were various challengers until Arbitron finally went belly up and sold out to Nielsen. In the magazine field, the leading service for a long time was Starch, which focused on primary audiences, then along came the Nielsen Media Service --NMS--- and Simmons and the Brand Rating Index---BRI--- and later The Target GroupIndex--TGI--and still later MRI, which eventually absorbed Simmons.
The reasons for all of these services were mainly that many approached the problem in different ways. BRI, TGI and MRI used the recent reading method for magazines while NMS and Simmons used "through-the book" ---a visual recognition methodology. The answers were very different. In TV we had meters vs diaries---again different answers and radio morphed from telephone coincidentals to diaries to PPMs over time.
Now we have a virtual monopoly with Nielsen---which everyone accepted without protest for years, and despite various efforts to unseat Nielsen none of the would be rivals has been able to show that Nielsen is substantially wrong in its findings. My take is this; if one or more of the supposed alternative sources offers a cheaper way to handle one of the emerging platforms, Nielsen will probably buy it---rather than being unseated by it.
Kind of surprised no one has mentioned Comscore/Rentrak for TV measurment. Beats Nielsen hands down. Is no one aware of Comscore? really? If you buy TV, you should.
Even today, you cannot replace predictability and stability in measurement unless you are ready to risk short term disruption. That has has happened before and led to improvement. But multiple currencies are expensive, problematic and unnecessary. You won't make more money with more currencies. All you will have is more expense. Research suppliers, not users, love the idea. Who is going to analyze all that data? Depleted research teams of today have been slashed for short term cost savings. History shows that media businesses with multiple currencies settle on just one. Perhaps Nielsen has finally gone too far and clients are ready for an alternative. I suspect it is closer than current Nielsen management imagines. Many years ago I looked at SMI's ad data, pooled from many large agencies, as an idea worth exploring as a model of collaboration for video measurement. With permission from Turner Sales Head David Levy I proposed sharing our digital data with anyone who was willing to share with us. The end game would have been a data pool, sourced from many networks/companies, that could be shared and combined with any other data, including Nielsen. Layer solid user friendly analytic software (Never a Nielsen strength) on top and you have a money saving alternative to the current system. A similar idea was proposed by CIMM. The industry reaction at the time was deafening silence. Maybe it is ready for an SMI model today? If you're going to tear the house down, the responsible path is to make sure the alternative is better in practice. Railing against the current system won't change much unless there is an alternative that is better, or faster or cheaper.
My two cents. Nice to be remembered by folks I admire, like Dave Morgan.
Everybody knows them. But few measurement experts believe they are an improvement. Digital has led the way to understanding the need for persons based measurement. Comscore has demos, but Comscore TV measurement is not persons based as it is in digital. Check it out.
Michael, the set top box --STB---approach gives you a much larger sample--which is fine for rating stability when you are measuring very low rated shows or slicing and dicing the information in a "granular" fashion. But are STB cable homes representative of all homes? I think not. And while Nielsen requires that its individual panel members indicate whether they are watching whenever a channel is selected, the set-top-boxes can not tell you who is "watching". That's a very big limitation.
Very good point Ed about the fact that whle we always had market leaders in each measufrment channel, we typicallly had competitors nipping at their heels pretty regularly.
Michael, Good point. I should have mentioned Comscore/Retnrak. They are certainly players in TV, though heaviest in local and I was a bit overly focused on natoinal. It will be interesting to see if they can make a run at building a full natoinal panel, or if they attack on the edges and push hardes to win in streaming.
Great persoective Jack; super helpful context. I do think that Nielsen has a shot to win back a lot of trust if they can nail NielsenOne, though htey would porbably be in a stonger positoin if they had maintained ownership and full control of the Gracenote data.
Also, sorry for the typ on your name :-)
One issue in media measurement globally, is that (in general) the owner of the media property wants to be a subscriber to a system that has coverage, mass, inclusivity, accuracy, stability and speed.
Oh, and of course at the lowest possible price. That generally leads to a lot of 'shopping around'.
The Australian and UK models are well worth looking at, where an industry body (JIC) is set up to run the system.
Here in AU we have OzTAM that is doing some really good and fucky things. While OzTAM is financially owned by the commercial FTA broadcasters, it is run by top quality researchers. The JIC includes the commercial FTA broadcasters, the public FTA broadcasters, STV entities, the MFA (ad agenices) and the AANA (advertisers). There is an independent auditor of the panel and 4-weekly reporting of key panel data. The TechCom meets monthly and discuss issues, improvements, developments, trends etc.
Here's a slightly different take. I worry that we're distracted trying to solve for 2012 research needs- fixing linear TV currency measurement- when what we really need to be focused on is a current state/future state integrated planning/activation/measurement research infrastructure. That's what media companies need to optimize yield of all their impressions; that's what agencies need to insure they're spending their advertiser's media dollars wisely. Linear TV measurement is a necessary component- there will still be linear TV ads bought on broad age/sex demos in 10 years, but it's clearly shrinking in terms of importance. And building that infrastructure is hard- requires panels, first party data from walled gardens, common target audience definitions across platform, ability to integrate RF forecasts for linear and addressable media, segmentation schema that reflects the way that media is consumed. Hopefully the WFA/ANA is on the path to solve for this.
So true Howard. This is not the time to solve last decade's problems. Doing the multi-channel digital media measurement right is giong to take a lot of work.
The specs for a new TV/video measurement that covers all of the ways that people access content are pretty obvious. First, you need a panel that is large enough to measure the smaller audience platforms with some degree of stability. That means a lot more than the 100,000 people in Nielsen's national people meter panel---maybe double or triple that number. Second, the service must be able to monitor all of the screens, including out-of-home usage. Third, the service must also be able to determine what specific content---programs or ad messages---was on each screen for whatever length of time it was active. Fourth and very important, panel members must dientify themselves as "watching" program content whenever a channel is accessed via any device in any location. We can't assume viewing. Fifth, and this is critical, some sort of attentiveness measurment is required for all devices in all locations that are used by panel members. This measurement should determine whether a claimed "program viewer" watched content of any kind---including commercials on a second by second basis.
I believe that such a system is within reach and it will eliminate the "walled garden" issue----it doesn't matter whether a given service opts to cooperate---- and solves the critical flaw in current "audience" measurement---namely that we are measuring set usage not second by second "viewing" ---and this produces extremely inflationary findings for commercials in particular.
But who pays for this new service? If it's mainly the TV ad sellers ( 75-80% ) and, to a much lesser degree the ad agencies---but not the advertisers and OTT/tech folks, we've got a problem. Why would a TV network sponsor a service that tells advertisers that only a fraction of its average minute "audience" watches their commercials?
It's fine for everyone to sit back and ponder ways to solve the "cross platform" measurement "problem". But that part is fairly easy. The hard part is who leads the charge and who pays?
Spot on Ed.
There are many people and entities on the sidelines expressing their hunger for change and improvement, but they seem to lose their appetite when it comes to putting their hands into their back pockets.
I think I'd be happy with a starter being robust internet measurement on an audience basis (i.e. people) rather than the still prevalent traffic basis.
Fabulous comment Ed. You nailed it.
Media companies pay for data that will help them raise revenue. More data and nice to have visions that don't do that won't succeed.
Jack, for sure, any new offerings will need to be funded by the media owners/sellers, since they're the only ones will real budgets, and vision won't be enough. If it can't help them sell more ads for more money than they can with the status quo, it won't be funded.