Originally, I was going to title this post “Universe,” because that’s what it’s really about. But I decided it might seem derivative of a column we’ve already published -- “Universal Truth” -- that kicked off what is becoming a series of posts examining a single subject. You know, the universe. Especially the parts we can see, how we organize them and how they define the way we think, how we feel and what we do about, well, everything.
Obviously, it’s a pretty big subject, so let’s narrow it down. For this series, we’ll focus primarily on the Media Universe -- especially the parts we cannot see, but which influence the way we think, feel and do things.
Like the greater universe it exists in, the Media Universe is a function of how people define it.
advertisement
advertisement
A few years ago, astrophysicists completed a project to measure and weigh the physical universe, and they determined that 99% of it is something they could not actually see and knew nothing about -- dark matter. I’ve come to believe that the Media Universe is just like that -- that there are dimensions of it that we cannot see and know little or nothing about.
The reason that’s important is that Media Universes aren’t just scientific theory. They are the basis of real markets that exchange trillions of dollars based on certain assumptions about their dimensions -- especially how big they are and what they are composed of. Especially the advertising marketplace, where not just the supply and demand of media are determined by its universe estimates, but all of our most important theories of media planning. Everything from reach and frequency to recency to attribution and even mix models depends on them.
The best example I can point to is what has affectionately become known as the “Nielsen Universe.”
Originally, the Nielsen universe was just a euphemism for saying the “Television Universe,” because it was Nielsen’s methods and data that determined the size of the TV marketplace people traded in. These days, the Nielsen Universe has grown to be much bigger than TV. And its latest deal with a mobile user identity startup PushSpring is an effort to capture a big part of the Mobile Universe.
There are many reasons why that’s important, but one of them is the power that Nielsen has to define the universe that powers advertising and other media markets. If you are part of the Nielsen Universe, there’s a good chance you can compete in it. If you’re not, it’s very difficult. That’s a lot of market power, but inherently, the marketplace has given Nielsen that power, because of one of the things the advertising marketplace loves most -- brands.
The Nielsen brand is so powerful in terms of shaping the industry’s belief systems about the size and composition of the universes it trades in that it has become de facto currency for much of the media marketplace. That’s mostly true for television, and increasingly true for digital. And after today, a little more so for mobile specifically.
The problem is that Nielsen doesn’t actually know the size and composition of any of the universes it measures. Just the parts it can measure.
In what is considered one of the most structured parts of the media universe -- national television -- Nielsen only actually measures viewing to about 70% of it, according to various industry experts. A lot of it gets down to how you device television, and even how you define television viewing, but for better or worse, the Nielsen Universe has become a proxy for it. That’s better for big TV companies that are part of it. It’s worse for the ones that are not.
The reason for calling this post “Coverage,” is a Media Universe is only as good as the coverage you have of it. And as good or bad as Nielsen’s coverage of the Television Universe is, its coverage of the rest of the Media Universe is even weaker.
And the big question for the industry is how much coverage does it take before efficient markets emerge around it. It has worked for TV. It has worked to minimally viable degrees for most other media. It hasn’t been working that well for mobile, which is one of the reasons why Nielsen is focusing so much on it.
Put all the side issues -- viewability, human beings, fraud, user experience, etc. -- aside for a moment. If you ask me, the biggest problem with the mobile media marketplace is that nobody knows how big it actually is. There are lots of numbers thrown around about handsets, installs, even time spent, but from a classic media planning universe point-of-view, nobody knows what its dimensions are or what it is composed of.
Now Nielsen, through its eXelate unit, is acquiring billions of mobile user IDs. If I do that math right, that makes it even bigger and more significant than the user ID deal it did with Facebook a few years ago that became the basis for much of its current Digital Universe.
If you’re like me, you probably never heard of PushSpring until today, and I have no way of knowing exactly how representative they are of the Mobile Universe, but I’ve asked some people at eXelate and they said they are going to get back to me with a couple of numbers. One is their best guess for the size of the Mobile Universe -- at least the universe of mobile user IDs -- is. Another is what Nielsen eXelate’s coverage of that is.
The reason that’s important is because Nielsen is competing with other powerful players who want to define the universe, and leverage the market power associated with it. We don’t call them universes -- we call them “walled gardens,” and they are companies like Google, Facebook, Apple, Amazon, etc. that have amassed control over a sizeable part of the universe of user IDs and they are leveraging them in ways that are making them the new most powerful market forces in the advertising industry. All because they can define -- and to some extent control -- parts of a universe that brands and agencies want in on.
How the industry ultimately works with those platforms -- those walled gardens, or aggregators like Nielsen eXelate, or even the next new startup like PushSpring -- ultimately will be a function of many business factors. But the truth is that as more and more of them emerge, it will get harder to understand what the actual universe is. And without knowing that, how can you know what share of it you have?
In other words, to paraphrase Dan Rather, “What’s the reach, Kenneth.”
Joe -
Once again - an insightful thought provoking article.
Well done.
Joe,
Excellent essay. I spent a year working for Arbitron before it was bought by Nielsen. Nielsen needed the Portabale People Meeter technologny, which it disparaged until the purchase. I'd say that the Portable People Meter accounts for about 90% of what Nielsen misses with its current technology. The Arbitron plan for the Portable People Meter included adaptors to allow it to measure media on smartphones and computers and tablets. The Achilles Heel, as I saw it, is compliance. The more complicated the measurement method becomes the more difficult it is to enforce compliance by the panels.
All Good Wishes,
Ted.
Joe, I have a difficult time accepting the notion that Nielsen only measures 70% of TV viewing. If that were true, the average person spends 8 hours a day watching TV, not 5.5 as is being reported currently. Add to that 2.5 hours for radio, another hour for print media plus the digital stuff and you have much too much media activity per person to be credible. There's been a lot of speculation about the "long tail" of unreported, not unmeasured, TV viewing, mainly for digital cable channels with very small audiences. If these are the basis for the 30% "unmeasured" figure, I just can't buy it. The combined viewing of all of the ABC, CBS and NBC programs in all dayparts is only about 15-16% of all viewing. How could the "long tail" channels amount to 30%?
Of course there are other viewing venues---SVOD, YouTube, etc. as well as out-of-home video/TV, which may be counted in the 30% figure. And nobody is against getting a proper fix on the viewership on such "platforms"---but how does one measure said "viewing"? It is claimed that the PPMs take care of this, however, the core assumption, namely that if the electronic device can "hear" an audio signal from a TV channel that its wearer/carrier is automatically "viewing", has never been validated or even explored. At least the peoplemeter requires panel members to signify whether they are "viewing" at the outset of a set usage occasion or when the channel is changed. But not the PPMs. So, imagine a situation at a bar where a football game is on the tube and a mob of partons are guzzeling their brews, chatting, socializing, etc. and some, of course, are watching the TV. If they all wore ppms every one in the bar would be counted as a "viewer". Is that really true? If you were an advertiser,would you pay the same CPM or higher to "reach" such "viewers" as you pay for in-home audiences?
I'm all for measuring the heck out of everything, but isn't it about time for some validation of the research instruments and the underlying assumptions, not just the mechanics of sample/panel selection, data processing, weighting systems, etc?
Excellent.
A very thought provoking piece Joe.
I have had to wrestle with issue of defining a media universe many times - and it is darned tough. All we know is that it is somewhere between 0 and the market's population.
We use methods such as Establishment Surveys (ES) which traditionally have been used to redude the population count to just include those that fall within the 'media definition'. This is becoming increasingly problematic with the digital onslaught. One example is trying to quantify and measure the DVR 'sub-universe' of the TV universe. The thing is most people in the ES will know whether thay have a TV or not. Most will know whether it is digital or analogue.
But when you start probing about DVRs it becomes much murkier. For example they may have STB which also has a DVR thay they didn't know was included. Or they may have bought a new IPTV and not known that it had storage and DVR capability. It is quite common (around one-in-ten) that homes that say they do NOT have a DVR are found to have one of some sort (in an STB or IPTV) when the home joins the panel. Where do you set the universe penetration - the claimed ownership or do you adjust for the 'false-negatives' and set it higher.
But traditionally each medium has set its own cap on its Universe. For example, radio is most commonly People 10+. Readership is generally People 14+ or People 18+. TV and Internet is People 2+ (we have TV as People 0+ here in Australia with the parent being the proxt button-pusher for the infant).
So, while I agree with you Joe, I think TV has a lesser issue than other media, as its coverage is usually much broader.
I also think Ed is on the money that the TV coverage at 70% is too low - I'd guess high 80s.
And regarding the PPM I think that 'false positives' of 'viewing' in a bar just because the volume is up on the TV is less of an issue than compliance. For example, PPM carry rates tend to be lower first thing in the morning. I know that when I trialled them here in Australia my 'red crosses' where my self-completion diary was different to the meter was because I had fogotten to carry the meter - and a problem with personal meters is that once you lose that data there is no way to 'back-fill' it, unlike in a diary where you can go back and tick the box. In fact, the respondent also tends to filter-out background volumes simply because they can't recall it meaning that often diaries provide better listening behaviour!
Cheers.