Nielsen Strikes Smart TV Data Deal With Vizio
- by Wayne Friedman , April 26, 2022
Nielsen now has the rights from Vizio to integrate Inscape data in both its local and national audience-measurement solutions.
Inscape captures viewing data from millions of connected TV (CTV) devices -- including cable, set-top-box, DVR, and streaming platforms -- through smart TV sets via Automated Content Recognition (ACR) technology
Nielsen says its TV panels will be used to “validate” what is missing in Big Data sets. This includes over-the-air inclusion, “validation of diverse representation," determining whether a TV set is on or off, and persons-level measurement.
advertisement
advertisement
The company says this will significantly add to the launch of its new cross-media measurement system Nielsen One, the forthcoming single deduplicated measure that will start this September.
Previously, Nielsen made a licensing deal with Inscape for national addressable measurement, starting in 2021.
"captures viewing data ..." "Viewing"? Really?
Unless I am missing something, Inscape and various other video data tech companies capture content rendered, aka "viewable impressions" per MRC, on a screen using ACR. No persons survey or persons measurement is invloved whether in the room, pressing meter buttons, having eyes-on/ear-on or paying attention. ACR data is at best tuning data and I am sure Nielsen understands this restriction.
As The Attention Council reminds advertisers, without Eyes-On/Ears-On at minimum and preferably "attention" there can be no creative message effects. For Media Post, referencing "viewing" when it is not supporrted by the methodology is beyond disapointing and inappropriate. How Nielsen will use this ACR "tuning data" to enhnance its Nielsen One persons ratings is, I suggest, the real story.
Do Vizio and other smart TV buyers know their "viewing" data is monitored?
Tony I suspect that in the Nielsen panel they will homes with ACR and be able to derive some form of 'ACR viewers per household' for ACR content. They could then feed in the Vizio 'delivered' data and apply the 'ACR viewers per household' factor which (theoretically) should provide a more robust estimate and for a lot more reportable programmes.
Which raises a point, John. To what extent are any of these "big data" panels vetted as to what universe they represent and whether they are a truly random sample of that universe. How are homes selected to be in such panels and what are the cooperation---or opt-in--- rates? What about panel turnover and the addition of new members?In the case of a set-top-box panel to what extent does it represenyt all "pay TV" homes? Are "pay TV" homes typical of all homes? etc. etc. I suppose that all sorts of weighting systems will be employed to mash all of this together---then add the refinement of a viewer-per-set projection from the relativeley tiny people meter panel to produce individual program and commercial "viewing" estimates. But is this really a "measurement" or is it more akin to statistical guessing?
John & Ed:
Amen. At best a highly complex simulation but agree ultimately specious.
Ed I agree that there is not an overseer for ACR data, but I think a time will come when Smart TVs are virtually ubiquitous. If ACR also has universal standards I think the data returned would be pretty robust. There will be 'holes' in the data as I doubt that all content on all Smart TVs will be able to be identified, but we have always had that problem with, say, local broadcaster content.
If my prognostication is close to the money and there is a research body that can aggregate, say the top 10 ACR data sources, then we'd have a pretty good handle on what is being shown on televisions in the non-linear world.
The big "BUT" is that is still only content recognition and not content viewing - that is, it is not audience data.
But merging a large TV viewing behavioural panel with large volumes of ACR content has great promise if it is all done correctly. In essence the panel would be focussing on the characteristics of the viewers, while ACR would focus on the volume of content on the screen. That would allow 'cleaning' of the ACR data (i.e. reducing the volume by a 'known' volume of content that had no-one watching the screen - for example only watched half the program). It would then be able to report the content volume to total people audience (i.e. a bigger number), and then report on demographic data based on the panel's usage (i.e. a then smaller number).
John, for the next five years or so a fair amount of viewing---especially in older but heavy viewing homes will be on dumb sets. As a guess they may still account for 20-25% of all set usage in four or five years. They are not included in big data ACR panels. Also, are these randomly selected homes or do they simply collect as many as they can---subject to consent for monitoring the usage--then sample balance the whole thing so it looks like the total ACR popluation? Is it possible that heavy set usage homes are over represented in such panels?
The same questions apply to set-top-box panels? Since they are homes that have stuck with "pay TV" and these tend to include more older , heavy viewers than are in the total population, what adjustments, aside from sample balancing, are made to deal with this possibility?
Last but not least, if viewer-per-set factors---it's really viewers per program as there is no way to know who remains in the room or is actually watching on a minute by minute basis---- are projected from 20,000 or so people meter homes which have the button pressing option--- it seems that many "people meter" homes do not obtain viewing data---and such VPH factors are projected to set usage obtained form many millions of ACR sets and STB's are we really happy with so slender a VPH base to provide us with sex, age, income, falily size and other breaks. You can't use the household characteristics of the ACR/STB panels for this as you will get a huge distortion favoring younger-middle aged and upscale homes when the actual viewers tilt in the opposite direction.
Needless to say, the sellers who are driving this will be delighted to get the big"impression" numbers they need to sell advertising and raise their CPMs---and who can blame them---it's a smart move in my opinion. But one would think that advertisers --who in times past would have been very involved---might be concerned about where we are headed. Yet they are mostly silent. That's a dumb move---in my opinion.