Home >> Insights >> Thought Leadership >> Television audience measurement: How to optimize TV advertising campaigns

Television audience measurement:
How to measure and optimize TV advertising campaigns

Image of a girl watching TV and being exposed to Linear TV ads that are measured by Beatgrid the cross-media audience measurement and ad effectiveness currency

practices in using TV next-gen audience measurement solutions to optimize the quality of reach

Optimizing quality reach is about accurately tracking the TV audience across all sources of streamed media viewing. In other words, to single-mindedly track the ads and seamlessly de-duplicate the data.

To begin to do that effectively, one has to look beyond the confusing CTV platform acronyms and start to plan for CTV as a new audience for TV that currently has multiple buying points (it is not a single channel!). For the viewer, it’s a new, rich world of viewing options for TV-quality content right there on the TV screen.

For advertisers, important new sources of cross-channel audience measurement data are exposing that media planning for CTV, by devices, platforms, and acronyms are indeterminate and need to be audience-driven.

CTV has quickly evolved over the past years and will continue to do so as the TV audience fragments rapidly. To avoid CTV developing into an entropy zone for media planners, randomness and uncertainty must be addressed. Precision and clarity are essential pillars in developing determinate best practices for the holistic measurement of, not only, an ad campaign’s efficient 1+ reach, but also its quality, or effective, reach across all streamed content sources for the viewer.

Single-source measurement data has arrived at a timely time for brand advertisers who are now confronted with multiple buying points to execute a TV ad campaign without increases in media budget to cover all the bases to maintain reach. Single source data helps the process of clarification and demystification of CTV, making it less indeterminate for media planners.

Demystifying the “CTV” audience

To understand why this is so necessary to implement a better planning practice, we first need to demystify the CTV audience. To date, there has been a tendency to technically define CTV by how the content gets in front of the viewer or by the commercial model underpinning how the viewer accesses a program. It’s important to remember that for the individual TV viewer it’s all TV. The fragmenting TV audience results from the wider choice of TV content now increasingly available beyond linear TV.

Given all this entropy around the current fuzzy definition of CTV and the impact it will have if unchecked on reliable, holistic TV audience measurement, not only does OTT streamed media need to be seen as one and measured accordingly, but it also must be seamlessly measured with linear TV.

Let’s take a step back and recap on why it’s become so confusing with a plethora of CTV acronyms in wide parlance now (just to focus on the most common ones): 

Linear TV is TV content produced and curated for broadcast to a daily announced schedule and includes channels that have +1 channels that run an hour behind the original broadcast, and so are still scheduled. These channels are all delivered to the viewers’ TV screen from their preferred TV operator’s platform, using a variety of content delivery technologies and household set-top boxes (STBs).

But first things first, what is VOD (Video-On-Demand)? This is where the rubber hits the road for advertisers wanting to optimize their TV audience because real data now exists on cross-channel audience reach and incrementality. The focus on defining CTV by how the content gets in front of the viewer, on which device or platform it is viewed, or even what the commercial model is for the viewer to view the content must be left behind to move forward with the accurate measurement data now available.

We then have BVOD, Broadcast VOD, which is the main reason why better clarity is needed in measurement going forward. BVOD is scheduled broadcaster content that has passed its viewing date. This BVOD content can be watched either through the familiarity of the TV operator’s STB platform or by exiting that platform and using another remote control to view the same content but now streamed over the internet (“over-the-top”, OTT) of the main TV service providers platform.

However, Streamed OTT services are not accessed through the set-top box (STB) but through another device. That device can be the Smart TV itself with its selection of publisher’s channels nicely packaged up in a “TV app”, increasingly now known as DTC apps (direct-to-consumer). That device can be a separate internet-connected device such as a game console, a proprietary black box, or a stick offering aggregated TV content. It can also be a mobile device

Lastly, let’s talk about AVOD, an Ad-Supported VOD streaming video service, funded by advertising and has the same model as commercial TV. AVOD services are free to viewers and in the CTV context are accessed via TV apps, and FAST (Free Ad-Supported Television) platforms that offer a combination of linear and on-demand long-form, high-quality TV. 

What is Connected TV advertising?

In a nutshell, Connected TV advertising or CTV ads, is skippable and dynamic advertising seen on the viewer’s VOD service of choice. It allows for better custom audience targeting as it can be optimized for the right channels. A fundamental advantage of optimizing for CTV ads is that it has the capability to reach more targeted audiences than linear TV, allowing advertisers to reach TV viewers that they would otherwise not be able to reach with traditional TV commercials.

How are CTV ads bought, sold, and measured?

Ads on CTV are sold directly by the broadcaster or an OTT TV service provider. On the buy-side, ads are bought by advertisers and their media agencies either direct from the provider or through DSPs (sophisticated demand-side platforms that buy campaigns programmatically). This is an evolving landscape. Accurate audience data for media planning is now an option to better understand CTV Reach & Frequency and CTV incremental reach.

How does the CTV landscape look?

Measurement data is confirming that the harder-to-reach TV audience of 18-35s that is moving away from Linear TV fastest can be reached through consistent inclusion of CTV on the media plan.

It is a fast-evolving ecosystem of competing publishers. Major publishers and broadcasters seek pervasive distribution of their content on all (audience) relevant OTT TV services and devices. That includes Smart TV itself with its own selection of publisher’s channels packaged up in a ‘TV app’ hub. The access device can also be a separate internet-connected device such as a game console, a proprietary black box, or a stick offering aggregated TV content. It can also be a mobile device. The landscape is further evolving with the arrival of high-quality, premium video inventory direct from Film Studios and Production Houses. TV and mobile apps are used to engage directly with their viewer base. These apps are known as DTC apps (direct-to-consumer). 

Image of today's fragmented TV advertising landscape

“There are audience measurement gaps to be filled in this fusion of data approach and it is likely that a good percentage of viewing remains unattributed as a result, maybe as high as 20% I overheard once?”

How CTV enables creative ad optimization

Bought programmatically, advertising can target key audience segments with the precision of online digital. For product advertising (not brand) this can be highly effective if the ads served are relevant and in context to the targeted viewer with clear calls to action. More precise measurement of optimal ad frequency across multiple buying points for an advertiser also allows the effective optimization of the impact of individual creatives in a campaign. Which ads are most effective in moving brand KPIs? When has an ad been seen too many times, on average, across all viewing sources to have any effect?

Household-level audience measurement

Linear TV advertising is officially measured, for trading purposes, using household-level viewing data paid for by the broadcasters. With the increase in the BVOD audience, device-based data is now also whenever anyone watches content on BVOD on other devices. This is done by the broadcasters tagging their BVOD services to generate BVOD viewing data on the number of devices in the home that are being used to watch BVOD services. This data does not capture the number of people watching or who they are.

The design of the next-generation TV audience measurement solution needs a mobile panel component.

Official viewing data is used by TV operators alongside their own proprietary STB data to fill some of these gaps and collaboration between the key players is increasing the amount of broadcaster data now becoming available to advertisers around BVOD.

The fusion of multiple sources of data does lead to a sense of progress in the quest for a cross-media measurement solution, and the industry is rightly pushing hard in this direction because advertisers are demanding it as the TV audience continues to fragment to OTT streaming, with AVOD and FAST at the forefront. This is accelerating the measurement challenge with each new publisher entering the game with their high-quality TV inventory. And it is fair to assume that there’s no single source in town that has fully indexed all this growing AVOD inventory for advertisers. To fill these increasingly extended measurement gaps, it makes sense to focus on tracking the actual ads themselves and to use methodologies that have first-party data across the widest number of channels possible.

The missing piece of the cross-media (TV) measurement jigsaw

The piece of the cross-media data design jigsaw that is missing for TV in this evolving design is passive mobile panel data. A nationally representative mobile panel (matched against the region’s general population representation) goes a long way towards addressing these measurement gaps in a way not possible by other methodologies. This is the destined way to pave the road for the cross-media audience measurement industry. Collecting high-quality, single-source data at a person level. 

This simple and effective measurement methodology is crucially device-agnostic, and therefore moves the focus of measurement to the audience itself, which can then be segmented for tracking alongside linear TV.

The beauty is that this is a methodology that is widely used now and for advertisers who want to get on with serious ad measurement across TV, and there’s no need to wait. 

It has been proven that single-source footfall attribution data can pinpoint whether a person has been exposed, or not, to the campaign when visiting the store and determine if they are, therefore, attributable to visitation uplift analytics.

In this context, a campaign’s measured effectiveness in driving store visitation, and particularly visitation uplift, can be designated as mid-funnel attribution because this insight does not allow the retailer to attribute an “uplift visit” to conversion further down the to the bottom of the funnel.

Deliver on efficiency and effectiveness of media spend across TV

I re-read a great book by Robert Pirsig and with full accreditation to him, I can envisage a shorter book, Zen and the Art of TV Audience Measurement. His book is, of course, on the Art of Motorcycle Maintenance, an inquiry into values in which he explores his concept of quality.

The underlying form of a TV audience measurement solution is what powers the effectiveness engine. Every detail in methodology design makes a difference in the accuracy of performance and results and needs quality data to deliver on efficiency in optimizing TV for quality reach, extending unduplicated reach with real first-party data on incrementality, and then using the same data to drive attribution (especially in campaign effectiveness in brand lift measurement).

Referencing again Pirsig’s enjoyable investigation of values, the value in demystifying the CTV space for advertisers, and reducing the state of entropy around it will be the start of upgrading the measurement of TV to a higher quality. The fewer data integrations involved means there are fewer hypotheses on how the data sets should be fused to get a reported result. Single source Cost-per-Reach (CPR) metrics are here, so single-source data is the way to go for optimizing budget allocation against a fragmenting TV viewing audience that is prone to go for the new channels found in the CTV realm.

Join the future of cross-media audience measurement and optimize your advertising with confidence and single-source transparency. There is no substitute for high-quality data because accuracy matters.

By Keith Johnson
By Keith Johnson

Beatgrid GM UK

Stay updated with Beatgrid's newsletter

Subscribe to our newsletter to learn about the latest news and industry trends.