The NHL is run by an old boys’ network where cronyism and lineage are more important than innovation and business results. How else do you explain the same old names being recycled again and again as new roles open up throughout the coaching and management ranks? The GM of the Vancouver Canucks, for example, is in year seven and over that time has put together one of the worst combined records of any team in the league, yet has a fresh endorsement from his boss. He’s an old school thinker disinterested in modern analytics and has signed some of the worst contracts in the NHL to go along with that abysmal record.
Few comments have captured the state of the NHL and its resistance to adopting a more modern approach than this tweet from mathematician and data visualization expert Micah McCurdy:
Of course, this isn’t a hockey blog and we’re here to talk media, right? But there are parallels in the worlds of sport and advertising that are hard to ignore. Hockey, like advertising, is at a crossroads where scratching past the surface is needed to bring clarity to complex narratives but a lot of decisions are still made based on gut rather than data.
Did you know, for example, that some of hockey’s most celebrated plays like blocked shots and hits are often associated with the worst teams and players in the league? Why? Well, the players that rack up these stats are often the ones that are always chasing or out of position, have the puck the least and are just generally being outplayed when they are on the ice.
Inspired by McCurdy, here are five questions a digital media planner may say “no” – or at least “not necessarily” – to based on the stories behind the numbers.
“Do these lousy view rates on YouTube mean my creative sucks?”
We like YouTube because of its strong reach and targeting, and skippable ads have a built in KPI around whether people choose to watch or not, but creative (and creative length) are just some of the factors that influence the view rates on YouTube.
A few years ago, and seemingly overnight, YouTube view rates started increasing dramatically even on some pretty ordinary campaigns. There are a few reasons why, but the main one is that, like Facebook, the algorithm simply got better at finding people who are more likely to watch videos. Today, targeting is one of the biggest influences on view rates, and in a counterintuitive way. Campaigns targeting bullseye audiences, essentially serving relevant messaging to the most relevant audience, often suffer the lowest view rates when compared to broader targeting because the platform has less flexibility in finding those power video watchers within smaller audiences.
Other factors include the devices ads are loading on, as more impressions serve on smart and connected TVs where people often aren’t ready – or don’t know how – to skip the ads resulting in strong “performance”. Finally, frequency and number of creatives play a big role in how quickly we see diminishing returns. No matter how impactful the messaging, someone who has watched your entire video may not choose to do so a third, fourth or fifth time which is why the 6 second bumper ad can be such an effective next touchpoint.
“What about this [really fun placement] on [really cool website]? I saw [really hip brand] on there – isn’t that a perfect fit for our audience?”
Everyone wants to see their name in bright lights and know where their ads are running, and there are a lot of great websites that are seemingly a perfect fit for a target audience you may be trying to connect with.
But when focusing on smaller or regional campaigns, scale can quickly become an issue with this approach. While a site may be a popular one among your audience, many sites have flat rates or enforce minimum spends, and once accounting for on target impressions in your market, CPMs can be in the hundreds or even thousands of dollars.
“What about this [really boring placement] on [really generic website]? I see [every brand] on there – isn’t that a perfect fit for our audience?”
On the flip side, choosing individual websites or apps that are too broad in their audience reach can cause the opposite issue. Yes, the target audience may frequent the website, but if it’s broadly appealing, likely everyone else does, as well. There may be a low cost to advertise on the website, but when you again consider the on target impressions and percentage of wastage, the CPMs can also be extremely high making it an inefficient way to reach the audience.
“Should we build an expandable banner?”
One of my favourite campaigns was in 2013 when our client was promoting a mobile app feature but their website wasn’t mobile-friendly. We can’t be talking about this innovative feature and dropping people on a pinch and zoom landing page, can we? So we pulled the page content into a rich media banner that people could interact with once they tapped to expand it. Makes sense, right?
But we know that engagement metrics are dubious at best and optimizing towards clicks or expansions, especially on the open web, can be a recipe for poor quality audiences and junky sites and apps. When you consider the cost of these ‘high impact’ ad units, and how many people in your audience actually view the expanded state, what you’re really paying for is a premium (often 5-10x) for the small mobile banners that sit at the bottom of apps and are generally too small to be effective. Interactive ad units can be a great way to stand out but it’s important to find ways to deliver them without requiring the audience to do something they didn’t go there to do.
“What about influencers?”
There’s a place for influencer marketing and some brands have seen exceptional success with it, but it can be an awkward fit for advertising campaigns that often aren’t squeezing enough juice out of existing channels.
Every brand should be setting aside budget to try new things and this an opportunity that’s always worth exploring, but with a measurement framework that will yield clear results, and again with a critical look at the number of impressions that will be in target. Many ‘local’ influencers can have a surprisingly large number of followers outside of where you may be targeting, and influencer strategies should look beyond just organic posts to paid distribution for greater audience control.
And the bonus round:
“Should we be using a QR code?”
No. Still no. Please, no.
Everyone wants to stand out in a cluttered medium, but good media teams dig deeper to find stories in the numbers and as a result might just save a million bucks.