Data never lies. Except when it does.
The importance of using data and magic to create a stronger brand.
We can collect data on most everything. Savvy marketers would be remiss not to capture all of the data available to them to measure the ROI of their campaigns. In theory, data helps you avoid the old concept of 50% of all advertising is wasted. You just don’t know which 50%.
Yet in his book Alchemy, Rory Sutherland suggests that digital advertising, while you can target people more accurately for less cost than traditional, it may actually be ineffective, particularly in the B2C world. Proctor and Gamble claimed to have reduced digital ad spend by $150 million without any impact on sales. B2B may be very different given the nature of products vs. the fickle consumer.
This highlights how you need to question assumptions about data and consider making bets to test them. One of the issues to consider with digital, and the inherent data available, is you risk doing what your competitors are doing because everyone has access to the same dashboards and trends.
If you follow the data alone, you eliminate the possibility for magic and serendipity. You need an element of unpredictability to stay a few steps ahead of your competitors. It’s not an ‘either or’, it’s ‘and’.
How data lies
We think data gives us objectivity. Yet objectivity is elusive as there is no way we can gather ALL the facts. There is more to know than you can ever learn. And what you think is right depends much on who you are and your experiences. Whether conscious or not, you inject bias into which facts you choose and what data you look at to support them. To actually be objective, you have to take yourself out of the equation. Not exactly practical.
Biases get coded into technology by humans doing the coding. That means the algorithms are never totally objective either. Nor is AI as it’s fed existing knowledge. We do hope the humans doing the coding have the best intentions.
“Data is always an abstraction of reality based on underlying assumptions as to how to categorize the unstructured phenomena of the real world.”
- Clayton M. Christensen, Competing Against Luck
We pay attention to data that confirms our assumptions. We suppress data that contradicts them. We cherry pick which data we show in our presentations to back up our actions to make us look smarter.
We want to make decisions based on data because it gives us something to base our decisions on. If that decision is wrong, we can blame the data. Not our decision making.
But if we make decisions based on art, intuition and experience, it’s much harder to defend if it turns out to be wrong.
It takes a lot of bravery to act with conviction without data. You need a culture of trust and a track record of good decisions for this to work.
The world is unpredictable. Sometime timing is everything. Sometimes you get lucky. And there are times when you’ll be wrong. History is a poor guide for the future. There’s a reason you see the fine print in all financial service products “Past Performance Is No Guarantee of Future Results”. This applies to life too. A good decision 5 years ago may not be good today.
“How is it even possible to maintain a data-driven approach to a world that simply does not make sense?
This is the future of work skill, which no one wants to talk about: Knowing when to look beyond the data, because only human intuition, leaps of logic and sense-making can come to ‘an’ answer.”
All of this doesn’t mean data isn’t valuable.
Data is still valuable. It’s silly not to take advantage of what we can measure with our marketing: the clicks, the downloads, the length of time someone watches a video before they drop off, etc. All of this helps us identify what’s working and what’s not.
But you need a solid strategy and purpose. Don’t just grab a bunch of numbers and try sense making. Go into it recognizing that just because it’s data doesn’t mean there’s certainty. Make assumptions knowing that there are margins of error. That it may not be accurate. It gives you a map forward. Clues on what to try next.
Embrace uncertainty. Think in terms of probabilities.
Collect data you think is relevant and then act on it, and if it’s a dead end, then revise what you’re collecting and how. Look at how the data compares with your intuition. Would you make the same decision without data as you would with? Does the data change what you’d do next?
What things can you try to test the validity of the data? Can you find a trend or is it an anomaly?
Like anything in life, data is not black and white. It’s one data point or three about the effectiveness of your marketing (pun intended).
You want to be able to verify your data is accurate. But for this to be realistic, you should also be able to verify that it is wrong. If you can’t verify it, then it’s not scientific, and thus biased. This is the scientific concept of falsifiability, according to science philosopher Karl Popper.
To make better informed decisions that guide your brand, you need to rethink how you use data to make decisions. There’s a science and an art to marketing and building exceptional brands.
Those of us that hide behind spreadsheets scouring endless reams of data ignoring our humanity end up creating bland brands. Brands that lack the magic that makes them rise to the top.
I’ll leave you with my ‘scientific’ formula:
Combine 2 parts data with 1 part weirdness and a healthy dash of analysis. Mix thoroughly. Test and iterate.
