Football and Marketing Part 1: Data vs ‘The Eye Test’.

As someone who spends at least 8 hours a day doing marketing stuff, and at least 24 hours a day being a pretty obsessive football fan, there are a few parallels between the two that have cropped up for me. And with the World Cup coming up, it’s definitely a topical subject to write a bit about…

So, I’ll start off with how both football and marketing have thoroughly and rightly embraced the explosion of available data – perhaps both industries over-index it at times. I’ll also give a disclaimer: I’m most certainly not saying that we shouldn’t use data as marketers. We should just think about its context a little more thoroughly.

Let’s start with marketing and data
Over the last three decades, the hegemony of data has become all-powerful in the world of marketing. There isn’t a single digital agency under the sun that doesn’t have the phrase “data-driven” somewhere on its website or marketing bumph.

Although creativity still has a critical role to play in marketing, targeting and statistical optimisation often take precedence. The ability to run detailed and (at least somewhat) rigorous creative testing means that, quite sensibly, we’ll nearly always follow the numbers rather than our gut feeling or subjective opinion on what works best.

At this point, I can’t count the number of times I’ve seen my preferred creative perform less well than something, well, less good. I’ve seen a Facebook Ad with a typo outperform every other ad in a high-volume account. I still can’t explain (without the use of conspiracy theory) why the initial iterations of Responsive Display Ads beat really well-designed HTML5 creative in just about every split test I ran.

Don’t get me wrong: I love data. I can’t design for the life of me. I’m a numbers nerd, which is probably why I’ve managed to build a career in digital marketing. I believe that the availability of data and self-serve digital ad platforms has had (or did have) a somewhat equalising force in business by enabling smaller businesses to be able to invest in and measure effective advertising – which was much harder to do 30 or 40 years ago.

But I do often think we’ve become over-reliant on data within our industry – and I also think that the ongoing changes to how online tracking works are going to leave many of us a little stymied. Not only are we going to lose the clarity of data we’ve become so accustomed to, but we’re also beginning to see that how people buy things is changing, and the ability to accurately separate correlation and causation is less and less simple.

Now for footy
Football (and sports in general) has undergone a similar statistical renaissance over the last couple of decades. Technology has enabled clubs to monitor more and more metrics about teams and players, which inform more complex strategies and ways of setting teams up. And spectators too are becoming more and more familiar with terms like xG, shot-creating actions and field tilt which would never have been thought about in the late 90s.

But here’s the thing: for all the data, all the sophistication of gegenpress or false 9s or positional play, football is still one of the sports in which underdogs win the most. Around 50% of games end up with odds being upset – compared to a range of around 35% for American Football (or padded rugby, if you will).

There are a couple of reasons for this. Firstly, it’s a low-scoring sport, which means that there’s a higher probability of anomalous outcomes. And secondly, we play it with our feet, whereas most games are played with our hands, and not many of us are more coordinated with the former than the latter.

And that’s where I see a correlation between football and digital. There are players who we watch on YouTube who look outstanding. Silkily skilled, both footed, can ping a ball top bins from 30 yards. And then they sign for a club and they entirely flop.

Here’s the thing: all the xG and xA metrics in the world can’t capture the fact that a player gets really homesick and hates the weather and the food in their new hometown. Just as all the data in the world can imply that a marketing decision is the right decision to make – only for it to not work out.

Sometimes the ‘eye test’, or the gut feel, are actually more important than the ad platforms or the analysts would have us believe. Because we can’t measure everything – and not everything measurable can be a simple statistic. So sometimes, things that we know are good, that are working well, that are doing something beneficial one way or another, won’t get picked up through our attribution or in direct contribution to revenue or however we’re measuring things. But we’ll know they’re helping. A bit like that midfielder who always runs 11km a game and is everywhere on the pitch but only gets a goal a season…

Balancing which numbers we take into account, being aware of data and metrics that are misleading, and sometimes (but not usually) ignoring the data and simply going with what our years of experience tell us can sometimes be right. That might simply mean that the data you have is insufficient I suppose – but it also ties into the fact that both football and marketing are highly unpredictable and chaotic activities, with random factors tying into who wins when.

For all our use of data – and believe me, I can’t get enough of the stuff – there are times when cutting through the noise and focusing on core principles is what we need to focus on. Just as there are times when Man City will play a bunch of wonderfully architected but ultimately fruitless football, but end up winning in the 89th minute 1-0 from an Ederson long ball over the top.

Next up: Football and Attribution Modelling and Why Gary Lineker is like Branded Search