Let’s measure up, not down
In the debates about cross-video campaign measurement, don’t forget what you can’t see on a spreadsheet, writes Thinkbox’s Matt Hill.
It’s a golden era for opinions about media measurement. Dinner parties sizzle with talk of minimum standards, school gates hum with debate over incremental reach, football stadia rock with songs of data auditing.
Maybe not yet, but in our corner of the world, campaign measurement is one of the hottest topics on media’s table — and rightly so. The fragmentation of media combined with walled gardens and return path device data have all contributed to the hot mess of audience measurement we now face.
Understandably advertisers have had enough of it. I don’t blame them. If I were in their shoes, I would also like to know the simple, topline statistics revealing how many people have seen my video ad campaign, where the overlap is, and where the unwanted frequency is. Am I paying to reach audiences I’ve already reached?
Unfortunately, whilst topline statistics may be simple in their final form, getting to an accurate number is hard, getting to a meaningful number is even harder, and getting to a number that everyone agrees on is, well, approaching impossible.
Phil Miles recently wrote an article for The Media Leader outlining YouTube’s perspective on the principles that should underpin cross-media measurement. It’s not a perspective everyone shares, and so is a useful place to bounce off to outline why different perspectives on measurement make it such a tough nut to crack.
Why setting a high bar matters
Take the definition of a view. YouTube advocates the use of Media Rating Council definitions as the base measure for video campaign reach measurement. That means an ad is on 100% of the screen for two tantalising seconds.
It’s intriguing that YouTube would opt for this. Of the many video platforms out there, they have one of the strongest products for ad view-through and one of the fairer charging models. If your YouTube ad is skipped, then you don’t pay for it (although I understand this can vary depending on what ad product you’re buying). The very low threshold of the MRC standard is much better suited to Instagram or TikTok.
One reason YouTube might prefer the low MRC standard for a view is that it makes its overall reach much higher than if it only included completed views, as Ebiquity has shown. And, as advertisers only pay YouTube for completed views, this approach looks like better value for money.
TV broadcasters do the opposite. They set the bar at the highest level — CFlight Broadcaster Video on Demand (BVOD) impressions only go into the model if they hit the 100% completed threshold.
This isn’t simply selfless, altruistic behavior by the broadcasters. They do this as they have a very, very high completion rate (around 97%), so it’s in their interests to set the bar high. In theory, they could take the same path as YouTube: only charge for 100% viewed ads but count everything based on MRC standards. This way they could throw in ads that have been watched for two seconds or fast-forwarded and add a nice extra chunk of ‘free reach’ to a campaign.
Reach needs context
But let’s not over-theorise. The point is that different media owners are going to have different perspectives on the measurement that best suits their business model.
A consequence of this is advertisers wanting the ability to pick and choose how they define a view. They are, after all, innately media neutral in their outlook, and this is something ISBA has indicated as the solution project Origin would adopt.
This is an understandable reaction, even if it’s hard on media owners who choose not to count everything all the way down to the MRC standard (or below). But it’s a scary solution when you consider another contentious point in Miles’s article — that the content environment is not something that should constitute a part of campaign measurement.
Regardless of how you count it, not all reach is born equal. The crucial qualitative differences between media are absent from reach curves. Plotted on the same chart, 7.9 million watching Andrew Tate on YouTube is the same as 7.9 million watching Channel 4’s The Great British Bake-Off. But that’s where the similarities end. Reach alone is one-dimensional and misleading without context.
I could write a tome about the hidden value of TV advertising that’s invisible on a reach curve, but will content myself here with some highlights to consider — high trust, premium content, sound on, big screen viewing in a relaxed setting, regulation, pre-vetting, high value exchange, shared viewing, costly signalling…
To ignore all this and focus on reach in isolation won’t lead to the best decisions — let alone if that reach includes everything from a glimpse to a completed view.
It once again highlights how important planners are to make sense of what is being measured — especially if the measurement is one-dimensional.
Cheap reach can be easily bought, but without the guidance of a seasoned, gifted planner, it’s easy to find yourself in an efficiency spiral of race-to-the-bottom planning. By focusing on the wrong metrics – like low CPT, or cost per start — you can wind up buying the lowest quality ad exposures.
It’s reach certainly, and it may in theory look like good value for money, but its actual effectiveness is going to be limited.
TV has great reach, but its quantitative curves shimmer with qualitative magic. In the debate about video investment — and measurement — don’t ignore what you can’t see on a spreadsheet.
Matt Hill is director of research and planning at Thinkbox, the marketing body for UK TV advertising.