In 2012, Red Bull attempted to bring its slogan “Red Bull Gives You Wings” to life. An estimated eight million people watched live on YouTube as skydiver Felix Baumgartner became the first person to break the sound barrier in free fall, jumping 39 kilometres to earth from a stratospheric balloon as the culmination of Red Bull Stratos, a seven-year brand-sponsored project.
The Stratos campaign sent Red Bull’s social media metrics sky-high. According to company stats, the event generated approximately 60 million interactions across all social media channels, and inspired almost one percent of all tweets sent on the day of the jump. And there were indications of a longer-lasting impact: Social media discussions about “drinking or needing” Red Bull were up seven percent three months after the campaign ended.
But all that buzz didn’t come cheap. Media estimates of Stratos’ cost have run in excess of US$50 million. And although the project is generally considered a big win for Red Bull, the standard methods companies use to measure ROI provide no direct way to trace how online chatter converted to real-world revenue. It is one thing to share or click a link from Twitter or Facebook, quite another to walk into a shop and buy a product. The familiar marketing attribution problem – identifying a set of user actions that contribute in some manner to a desired outcome such as a purchase – is getting even more complex in the digital world. The most common method for attribution, the so-called “last click”, grants full credit to the customer interaction immediately preceding the outcome, an easy but wholly inaccurate practice.
It’s time we started treating social media like any other marketing channel by assessing its impact in terms of genuine brand engagement and revenue, rather than solely relying on more ambiguous stats such as clicks and likes.
My thoughts on this matter echo those of my colleague David Dubois, who recently wrote on INSEAD Knowledge, “What businesses need to do when measuring ROI is take into account both the positional and relational to reflect social media’s true added value.” Luckily, in the last few years social media listening tools have become available, giving companies a thorough overview of not only how many fans and followers they have online but also the range of reactions to brands on social networks.
These tools make it easy for companies to get a quantitative estimate of how social media users feel about their brand (and its competitors). In the industry, this metric is widely referred to as “sentiment”, and is calculated in its simplest form by assigning emotional values to the words used in brand-related social media posts. (E.g., the word “good” would indicate positive sentiment; “awesome” would register as strongly positive). This is not an exact science, to be sure, but it has been known to produce broadly accurate and actionable assessments.
Of course, gauging ROI is a complex endeavor and can never be reduced to a single metric. But sentiment is an important part of the package because it measures more than an audience’s size. As Dubois points out, a company would probably see better ROI from 1,000 passionate fans than from one million disengaged ones. Having a tool that is accurate in measuring the sentiment is a very important first step (‘Garbage in, garbage out’). Once we have measured the sentiment over time we could use this as one input variable – besides many other variables – in a classical marketing mix model (MMM).
The “Marketing Mix Model”
In recent years, MMM has emerged as the gold standard for evaluating marketing ROI. MMM employs regression modeling and historical performance-based data sets to gauge the business performance, and to determine the optimal use of a variety of marketing channels. The results are often surprising to companies that have grown used to buying eyeballs: Using MMM, Procter & Gamble found that BeingGirl.com, a community for girls sponsored by the Tampax and Always brands, was several times more effective at driving sales than a similarly priced television ad campaign.
By plugging sentiment into MMM, companies would finally be able to assess the impact of social media conversations on business performance. This would lead to a better understanding of how social fits alongside more established channels such as paid advertising in a successful media plan. We could measure the correlation between the sentiment index and potential outcome variables like sales or market share and also understand how it would interact with other media activities. For example, a Facebook campaign may or may not be more successful if preceded by a TV ad. The downside here is that such analysis can be time-consuming and requires collecting a vast amount of varied data (sometimes even triggered through test and learn experiments).
From Facebook to the Factory Floor
In addition to measuring ROI, the champions of social media listening use the tools to spot and address product issues before they go damagingly viral. When Kraft Foods noticed alarming words such as “cut” and “blood” were appearing in online posts about their brand, they dug deeper and found some customers were hurting themselves trying to open a recently redesigned salad dressing bottle. Kraft took quick action to change the bottle, halting the potential for this minor boo-boo to become a more serious wound to the firm’s reputation.
This is yet another example of the importance of integrating social media strategy into brand or business strategy. (I wrote extensively about this in a previous post). These days, what happens in cyberspace affects companies in tangible ways; strategic objectives often succeed or fail online. Abolishing silos facilitates the speed and responsiveness that these new conditions demand. And including social media in the marketing ROI package is a great way to bring the digital world back down to earth – or, as in the case of Red Bull, into the stratosphere.
Joerg Niessing is an Affiliate Professor of Marketing at INSEAD. You can follow Joerg on Twitter @JoergNiessing
Leave a Comment