Marketers describing themselves as ‘data-led’ or ‘data-driven’ is now so common as to almost be a cliche, and is just one of a myriad examples of the industry co-opting the vocabulary of the sciences (A/B testing, big data, regression analysis…). I’m not certain what underpins this shift, and in the spirit of pushing against it I’m not going to try to back up my thinking with any evidence right now, but it strikes me that one of two things is happening. Either we’re generally so bad with data that we’re in awe of it and we think this language therefore lends more weight to what we do; or we’re so uncomfortable with the inherent subjectivity of marketing that we cling to data as something objective, as though it somehow absolves us of responsibility – “the data made me do it!”

What does it mean to be ‘led’ by something?

Multiple studies into leadership have concluded that there has to be some element of deliberate action on the part of a leader, whether that’s in a traditional, managerial sense or through a more modern, empowering approach. As McKinsey put it, “leadership is something you do, not something you are”.

Data is merely a recording of what happened in the past. You can interpret it to identify what you think did or didn’t work in the past, and you can make predictions or decisions based on what you think it tells you, but the data itself can’t do any of that. That being the case, data can’t possibly ‘lead’. It isn’t sentient, it can’t act autonomously, and it can’t make any decisions in its own right.

Data isn’t conscious… but it can “lie”

Although it can’t lead, data can certainly mislead. There’s plenty of famous examples of this, both within the marketing industry and beyond. However, I’ll just explore one instance of how data, especially when it involves human beings, can hide the truth.

According to GWI data, UK consumers routinely say that sustainability is a key consideration in their buying decisions, with 48% of them saying they’d even pay more for a product for it to be eco-friendly:

If we take that data at face value then we’d expect to see eco-friendly businesses flourishing, and pretty much sharing revenues equally with their less environmentally-minded counterparts, but that’s not even remotely close to being the case. Patagonia are routinely held up as a paragon of green credentials and they certainly do make some serious money out of it – £1.3bn in 2023 in fact. Primark, who it’s fair to say fall at the other end of the eco-friendly spectrum, should therefore have seen similar revenues… but they actually drove £9bn in 2023. Clearly not all of that 48% were telling the truth!

There are unknown unknowns

The concept of unknown unknowns dates back to the 1950s but was most famously coined by Donald Rumsfeld during the Iraq War. It’s essentially an acknowledgement that we can never have all of the relevant information when making any decision so we need to be aware of that fact. However, not only have marketers seemingly forgotten this, there’s even a fair amount of wilful blindness where we pretend that the data we have gives us everything we need.

We need to remember that there will always be an element of selectivity in the data we have available, whether due to technical limitations, prioritisation or plain old human choice. None of this is necessarily a bad thing, but we do need to remember that what we’re not measuring might be just as important as what we are. For example, I worked with one client who saw email sign-ups drop off and a subsequent tailing-off in marketing performance. They looked at their data and decided that they needed to invest more money into marketing, trial new channels, and incentivise sign-ups in the short term, all at great expense. Only when none of this made an iota of difference did they look outside these metrics, where lo and behold they found that there was a dev issue which meant that the email sign-ups weren’t being passed into their database. In this instance, it was the data they didn’t even know they didn’t have that was the true cause of their issues.

Sometimes you have to wiggle when everyone else waggles

Now I’m not suggesting for a moment that we should stop using data, nor even that it shouldn’t be central to our decision-making. However, when everyone has access to pretty much the same raw data, it’s what we add in terms of the analysis and interpretation which empowers us to make strategic choices. We need to be influenced by the data, but not let it drive. If we don’t add that element of subjectivity then we will all make the same decisions, and do the same things. In strategic terms, we’ll converge on those data points, create red oceans and continuously erode margin in the pursuit of market share.

Why is this important?

Just as important as the insight we can glean from the data is also what it might not tell us, otherwise we risk over-optimising based on what we do know. Just like the small percentage of bees who ignore the waggle dance to explore new frontiers, we need to take occasional bold decisions which are not based on the data. If these pay-off, we can then harvest the data from this experimentation to optimise before once again venturing into the unknown.