What Marketing Can Learn from the Beautiful Game

Editor’s Note: In this article, our UK-based author is discussing football, known as soccer in a few parts of the world. We hope calling this out helps our readers navigate the article smoothly right from the first word, fostering a better understanding of data’s revolutionary power AND appreciation for the global game.

---

What marketing can learn from the beautiful game

Football. The most popular sport on the planet. For many, more than just a sport. An ethos, a philosophy, a way of life – defined by raw emotion, bitter rivalries and spectacles watched by billions.

Although marketing might not wield quite the same cultural influence as the beautiful game, the two phenomena share a surprising bond: they have both been revolutionized by data.

In the last few decades, data has gone from being a niche resource to an integral element of decision-making. Football clubs now analyze the most precise details of player performance, just as brand managers collect the most granular and real-time information about their customers. Both are increasingly convinced of the power of data to gain competitive advantage.

However, with the sheer quantity of data now available, it has become easier to simply present numbers than properly scrutinize them; a habit that presents problems for both football and marketing, given the growing appetite for data-driven decision making.

So what can football teach marketers about applying data, in a way that maximizes reward and avoids pitfalls?

Data tells you what not to do

We typically think data should tell us the best solution. In football terms, for example, the best formation to use or the best player to buy. But counterintuitively the real value of data is revealing the worst solution. In essence, the approach to avoid. Why is this?

On a practical level, hard data can rarely separate the best options. There are small tangible differences between top players – for instance two world class strikers tend to perform similarly in terms of goals and assists. Deciding who is better therefore relies on unquantifiable soft skills like leadership or mentality, areas where numbers are inherently less useful than personal judgement.

So the quantitative element of data means it can’t often distinguish between the best. Instead it’s more suited to revealing which options are worse.

This principle is particularly helpful for football transfers. In 2010, the data clearly showed that strikers Marouane Chamakh and Park Chu-young weren’t worth signing. They took shots from improbable positions, they contributed little from open play, and their expected goals were low. Arsene Wenger – then manager of Arsenal FC – took the gamble and signed both players. But in the end the data got it right. The strikers achieved an unimpressive record of 15 goals in 74 collective appearances for Arsenal. In other words, they should have been avoided.

There’s also a strategic reason why data is more valuable as a tool for elimination not optimization: it forces you to consider a wider set of options, freeing up more potential paths to success.

In the world of marketing, Audi Denmark used this counterintuitive idea to their advantage. Rather than defining and targeting the small group of consumers most likely to buy the brand – the standard approach for online ad campaigns – Audi defined those who rejected the brand and targeted everyone else. This meant they were going after a much broader target audience, all of whom had at least some propensity to buy the brand. The impact was stark: a significant improvement in campaign conversions, and a compelling case for using data to avoid the worst, not find the best.

A declining metric can be misleading

As a football manager, you’d probably worry if your star player was underperforming on a key measure. At least this was the case for Manchester United FC boss Sir Alex Ferguson, when he noticed defender Jaap Stam was making fewer tackles than in previous seasons. The data convinced the manager that Stam was past his prime, leading Ferguson to transfer him to Italian club Lazio.

But Ferguson had been deceived by data. Stam was making fewer tackles because he was making more interceptions and keeping the ball more as a result. In other words, the drop in tackles showed his performance was improving, not declining. After Stam’s successful stint at Lazio, Ferguson later recognised his mistake, admitting ‘without a question, I made a mistake there. Jaap Stam was the one.’

The story highlights an important point. Declining metrics aren’t always a bad thing, as they often indicate positive performance elsewhere. And the reverse is also true. An improving metric doesn’t necessarily signal success, as it may be driven by something else entirely. A finding that is just as applicable to marketing as it is to football.

Insiders at Ebay were convinced their Google search ads were increasing website traffic. After all, the data showed that the more money spent on these ads, the more people clicked on the link to the Ebay website. But once again, deceptive data masked the truth. The ads weren’t causing more people to click on the link – they were guiding a growing number of people who were already looking for the link. Ebay eventually switched off the advertising and found a negligible impact on traffic, soon realizing it was spending $20m on ads for nothing, and discovering the danger of misleading metrics.

You need multiple metrics to paint the full picture

It would be convenient if only one metric determined the success of football players. You could track it and know instantly how well each player was performing. The reality is that players need to succeed in multiple areas, and therefore must be judged on multiple metrics. And when a player’s role in the team changes, so do the numbers we should use to evaluate them.

This is best illustrated by the career of global superstar Cristiano Ronaldo. From his early days at Manchester United to his recent spell at Italian club Juventus, Ronaldo’s average dribbles per game declined from 5.8 to 3.1. He also took fewer touches per game (65 vs 51), and created fewer chances for his teammates (1.9 vs 1.2).

It seems like we’re looking at a decline in performance, but actually we’re looking at the wrong numbers. Now, at the ripe age of 38, Ronaldo is no longer the explosive winger who sprints past defenders and crosses the ball into the box. Instead, he is the veteran striker who scores decisive goals in the big games. And when judged on more relevant metrics, he is executing his new role brilliantly. During his time at Juventus, Ronaldo’s shot conversion stood at 17% – an all-time career high. Compared to his Manchester United peak, goals per game increased from 0.72 to 0.93 – a sign of his lethal effectiveness in front of goal.

Ronaldo’s career path shares similarities with advertising. No single metric determines the success of all ad campaigns, so what to focus on will depend on the chosen objectives. Take the famous Cadbury Gorilla ad. According to insiders, the campaign achieved a poor ‘persuasion’ score when it was pre-tested with consumers (i.e. tested before going live on air). But as is apparent to anyone who’s seen it, the ad makes no attempt to persuade consumers about new product benefits or new price offers. Instead, it’s designed for enjoyment and entertainment. Given the testing agency “had to recalibrate its own scale to measure how engaging and impactful it was”, it’s no wonder Cadbury decided to run the now legendary advert.