"On TV And Video" is a column exploring opportunities and challenges in advanced TV and video.
Today’s column is written by Ronan Higgins, CEO at TVadSync.
The MRC recently released final standards for cross-media audience measurement of TV and Video. These standards will require work for all players in the market to get a passing grade, including technology upgrades from several members of the supply chain.
At this point, online video is best off, connected TV (CTV) must grapple with some challenges, and linear faces the biggest uphill climb of all. For media buyers and publishers with multichannel strategies, getting to a fair apples-to-apples comparison won’t happen until all of these issues are addressed.
The MRC standards break down into four top-level requirements. In the first, viewable impressions require that there is 100% viewability on screen for two continuous seconds.
Additional standards include duration reporting – knowing how long someone watched depends on who is sharing that signal data, be it reporting from a broadcaster, measurement from a smart TV or data from a set-top box.
There is also a sophisticated invalid traffic (SIVT) filtration standard; before invalid traffic can be filtered, it has to be detected, and unlike on digital, there are no current norms in place to do that with TV and video.
The last standard is consideration for audio – where it can be measured. Audio might be the biggest technical issue with the plan, as many players along the supply chain don’t report it, or they report differently.
Smaller screens have smaller problems
For regular online video on small screens, most of these new MRC measurement requirements can be supported by viewability measurement providers such as Moat and DoubleVerify. There are complications in relation to VAST and native in-app ads, but VAST 4.0 should fix these issues when it is adopted at scale. Other than a few additional issues with using pixel trackers, which can’t filter SIVT, online video is in good shape.
For connected TV ads that show up on TVs, things get more complicated. CTV content can be viewed with a native app in the smart television or with an app on a device connected to the television, such as a Roku, Amazon Fire TV, Apple TV, PlayStation, Google Chromecast or even a cable/satellite set-top box. The viewability, view duration and SIVT requirements will be taken care of by similar tools used in online video, but the consideration for audio standard runs into problems.
The TV audio might be off, for example, and the connected device wouldn’t know it. It is technically possible for television sets to communicate their audio state via something called the HDMI-CEC protocol, but this has not been standardized and is not part of established audio measurement instrumentation. So at least for now, this pain point has not been solved.
The technical dilemma of linear
Applying the new MRC standards to linear TV will be the most difficult.
What is SIVT in the context of linear television? Invalid traffic is an internet term that encompasses bot fraud, click farms and the like. In the context of linear TV, the industry needs to know what is considered invalid traffic. If an ad is displayed on screen, but the viewer has left the room to make a cup of tea, is it invalid? If so we’re going to need more complex presence detection sensors.
These not-so-small issues prove just how different the standards are for the different screens under the prevailing logic. If linear ends up subject to similar viewing rules as digital, it’s possible that certain parties could set up “viewing farms” to game the system. (Sound familiar?)
I see several technical hurdles that will need to be overcome:
Video-based automatic content recognition (ACR) can, by design, measure a TV ad’s viewability and duration of time viewed. However, as a measurement instrument by itself, ACR will not be able to report on audio state.
Video ACR built into the television set could be paired with TV set audio state data (volume level, mute state), but this has not been implemented by any of the television brands and may require software or firmware updates for that data to be collected.
Audio ACR, by design, can record the audio state, but on its own cannot differentiate between the video impression duration and the audio play duration, as required by the MRC.
Time for an arms race
The MRC requirement for the consideration for audio standard basically calls into question measurement of CTV and linear TV ads by any system that is not embedded in the television. The MRC concedes, in coded language, that the consideration for audio standard must be measured “where feasible,” and audio duration reported separately “where this can be measured.” MRC also says “If an organization cannot measure audio this shall be disclosed.”
But in an open and competitive market, who is going to buy measurement from a vendor that can’t report audio? I see a bright future for in-TV ACR that can marry audio state with video duration.
Ultimately, we need common standards for all video across screens, and the MRC regulations are a good way for the conversation to progress. However, the problems will be painful to fix for everyone, and we need to make sure that buyers, sellers and middlemen of all kinds don’t game the system and erode any credibility that can be created with good standards.