Are We Measuring TV Ads With The Equivalent Of The Click-Through Rate?
When I started my career, the ad server was just taking hold, making digital advertising measureable. We were off to the races.
First, click-through rates: “This is amazing! People are clicking on banners.”
Then, post-click tracking: “Incredible. We are tracking online sales!”
Next, view-through impressions: “Hold on, we need to track people who aren’t clicking!”
Now, it’s multitouch attribution: “We can quantify every digital touch point in the consumer path.”
Those were the days, right? Everything was measurable, and everyone was confident in their decision-making – until they weren’t. The digital industry went from measuring things because they could to asking questions about whether they were measuring correctly, and it continues to iterate and adjust to this day.
Here We Are Again
Historically, behavioral measurement for a mass reach medium like TV was limited to direct-response marketers who sold everything via call center and could tie sales back to a unique phone number assigned to the ad placement. All other advertisers focused on evaluating their plans’ reach and frequency, and they used marketing-mix models to justify strategic investments.
Behavioral measurement and optimization is now expanding into the traditional linear television space. There a half-dozen or more tools, or you can DIY with Google Analytics and ad instance-level data, to measure the immediate response of users exposed to a television ad and the impact on an advertiser’s website.
In its simplest form, it is looking at minute-level data on a website, overlaying spot-level TV exposure and monitoring sessions, visits and maybe even registration metrics to gauge any increases above baseline levels. There are far more sophisticated approaches with ad tech platforms, but that is the general concept.
Is This The Right Metric To Use For Optimization Decision-Making?
I understand the appeal. You can look at it and say, “Look, this TV spot caused a spike in activity on my website.”
There is certainly value in this. A consumer who was exposed to an ad picked up a device, searched for and visited a website to learn more about a product. That’s fantastic, or at least it is interesting. But what should marketers do with this insight? Can they feel confident that this is a true representation of the impact of this investment, or even a true leading indicator?
Perhaps more importantly, what we learned in digital is that a click-based or immediate-response measure is about efficiency. However, marketers who need to leverage a mass medium like television are doing so for scale. Using an immediate-response metric as the sole means of evaluating this medium is counterintuitive.
We Need To Start Asking Ourselves Critical Questions Again
In my experience, the response rates usually fall between 0.05% and 0.15%. Do these numbers look familiar? They are very much akin to the banner click-through rates or even last-click activity.
So, what is going on with the other 99%-plus of the audience that is not picking up their device and immediately researching a product after watching an ad? Is that fractional percentage of the audience and their related behavior representative of success? Who are these people who are compelled to pick up their device and visit a website immediately after an ad? I am having trouble recalling a time when I did this personally.
Was it this single ad that triggered this behavior, or was this a function of greater frequency? Do these people look like the rest of my consumer base? Is this cohort of consumers converting at a similar rate? Is my goal to make consumers immediately rush to my website to learn about or purchase my product or service? Do I expect consumers to immediately disengage from what they are viewing to do so?
We need to be asking ourselves all of these questions to help us determine what to do with this metric.
I’m not suggesting that this measure isn’t valid; I’m suggesting that we take stock and proceed with caution on what we do with this metric. If we react to this data in a vacuum, we may optimize to false positives or negatives. These results can drive marketers’ strategic efforts to inventory that may seem efficient in the context of behavior occurring within minutes of a commercial, but an unintended consequence may be a campaign that cannot scale.
It is easy to focus on efficiency metrics with this approach, and low-cost spots will always win out in that view. Investment with larger reach and scale will rarely ever be justified in this view, as they will not drive enough consumers to “click” on your television commercial within a discrete window. Does that mean that the consumers who don’t exhibit this behavior are less valuable to a business?
We need to be mindful and thoughtful about how to use this data. Marketers should explore solutions that provide additional context around longer-term effects of their advertising investment and optimization opportunities to help triangulate the truth about how this investment impacts their business and bottom lines.
Let’s not fall into the same trap that we did with digital and put too much weight on a metric simply because it can be easily measured.