Everyone knows the Olympics aren’t just a series of competitions to anoint the best athletes, but also the World championships of brand advertising. On the Monday after the summer fortnight, people around the water cooler will be talking more about the ads than the events themselves.
But do all those hundreds of millions of dollars spent on ads really produce ROI for sponsors? Those answers have been historically -- and notoriously -- elusive for marketers. Social listening platforms can only share metrics on fluffy concepts like buzz and reach.
Today, innovative approaches are helping brands to directly measure how paid sponsorships can affect the bottom line, down to the dollar. The process starts with measuring revenue-driving conversations within target audiences on platforms like Twitter.
So how does it work?
Filtering out the spam, duplicates, bots, and other noise helps identify the intending consumer base. For example, a tweet that says something along the lines of “Wow that Michael Phelps is really handsome” does not signal that the author is going to spend hard-earned money to buy Under Armour products after watching an ad featuring the 18-time gold medalist during the Olympic broadcast. The only thing that tweet does signal is that the writer does in fact believe that Michael is a good-looking fellow. A share showing intent to actually open a wallet reads like “Can’t wait until I bring home my overtime check and get that new Under Armour bathing suit.” These sentiments correlate directly to revenue.
When advertisers create theme-specific advertising or air advertising during theme-specific programming, they are interested in how the intended target audience responds to that carefully crafted content. But it’s extremely difficult (if not impossible) to determine how certain demographics respond to an Olympic ad in terms of revenue in real time.
So instead of relying on cookie-cutter demographics (age, gender, geo, etc), advertisers can define audiences by how they define themselves and the content they are interested in consuming – by grouping them into interest buckets consumers opt into online. For example, if brands are interested in how sports fans responded to an ad they should look at the revenue-driving brand conversations of, say, all the people who follow at least one major professional sports team on Twitter. Now, instead of tracking a generic demographic concept like males ages 25-44, advertisers can zero in on performance within those who self-identify as sports fans.
Okay so then how might you determine whether someone is a sports fan? You could start by looking at the 3.29 million people who follow the Olympics on Twitter. Or the ~20 million people who follow NBA teams. You get the picture; sample size is not a problem.
Advertisers should then look at the increase in the number of revenue-driving brand conversations online within the target audience (in this case, sports fans). So let’s say in the 4 hours leading up to the Olympic telecast this sports fan audience wrote tweets showing intent to buy Under Armour bathing suits 1000 times. If, during the 4 hours of the Olympic telecast itself that same specific audience tweeted about buying the bathing suit 5000 times, one could reasonably conclude that an advertisement for Under Armour aired during the Olympic bloc upped interest 5x.
Now you might be thinking to yourself, “Wait, it’s the Olympics, there are going to be higher tweet volumes for hundreds of different topics. How can you know whether sports fans are really now more keen to buy UA suits after seeing the ad?”
The answer in this example lies in a comparison against the general number of people writing tweets showing intent to buy the suits. This is how you isolate for the impact of the campaign on sports fans. Let’s say that the number of total intending tweets about Under Armour bathing suits doubled from one 4-hour block described above to the other. In this case the ad targeting sports fans would be effective because that targeted sports fan audience grew 5x while the total general population audience only grew 2x during the same period.
The next step is to compare this 5x jump with similar historical performances within the same audience. Let’s say Under Armour also used the Super Bowl 6 months earlier as an opportunity to push the new suit. And let’s say that on average tweets showing intent to buy the suit during the 4 hours of the Super Bowl coverage within the sports fan audience grew 7.5x. One could therefore conclude that at 5x growth the ad for the bathing suit actually underperformed relative to the Super Bowl.
You can then step back and look, over time, at the impact of the Olympic advertisement on your targeted sports audience as it pertains to revenue, relative to expectations.