Small data: taming the big data beast
I was fortunate to be invited recently to hear Martin Lindstrom speak on what led him to conclude that small data (his coined phrase for small, human clues) is being wrongly set aside in favour of big data, and write his latest book on the subject.
Martin Lindstrom is a brand expert and head of the Lindstrom Company, who’s pioneered several ways of strengthening the way marketers go about their strategic processes, via his company and more broadly via several best-selling books.
It was an inspiring talk and I left it with vindicated advocacy for the power of emotional DNA and observations of human activity in crafting stories and activations that makes a brand great, which is something big data simply can’t do alone.
There were 3 key themes that drove the message home:
Big data matters though only when dancing with small data
Big data is one of the latest trends not only in marketing, but in life and universe. At its core, big data is the attempt to correlate different data collected from many sources into a view and according to big data hubris, a view that is unmatched in its accuracy and ability to predict outcomes, needs and wants.
But big data is machine language and this is where its limitations are found.
A well-known case study is in how Google, using big data, apparently predicted the timing and scale of the 2013 flu outbreak within days based – amongst other things – on what people were searching for.
At the time it seemed big data had arrived; with the data collection powers of the internet and then having those millions of pieces analysed and correlated, it seemed nothing was safe from targeting at an accuracy unheard of before.
The problem was that causation (why the data was created) was ignored by the rising of correlation (what was happening).
They got the scale wrong because while true that people were searching flu-related strings, they were not always sick, just curious (blame media). Google’s algorithms could never observe or talk to the human doing the search; all it had were keystrokes on which to base its outcomes and that’s where it was flawed.
Big data can’t measure causation, only correlation to reach its conclusion.
Of course, the collection and analysis of data always matters for marketers. It’s an important piece of the puzzle and because we have access to more data and information than ever before, so we should use that access. The danger exists in relying on the data alone where causation through observation is missing.
However, the best analysis and outcomes will depend on the application of both machine logic (big data) and intuition (small data).
Small data matters more against the backdrop of big data
Small data is about humans and while it’s humans that are buying your brand, with all of the emotional constructs that go along with being one, emotions will remain the most fundamental of concepts in marketing.
In preparing for his book, Lindstrom literally visited “thousands” of homes of people (their own environments) to interview and observe for small clues about why they think and act the way they do.
Some of the environmental clues Lindstrom cites from his research are provocative. For example, a huge bookshelf full of books could simply be over-compensation for a limited education. However someone with a large, colourful picture of themselves on the wall is very self-confident and I don’t think you can argue with that.
The point is that small data is almost about going back to marketing basics, from a certain point of view, before the data-driven world arrived, because emotional DNA doesn’t change. The observation of what causes people to do what they do is intrinsic to crafting a relevant marketing story.
Balance is needed between the information that can be gleaned from big data and the temptation to rely on that alone.
Small data provides the intuition and emotion that big data simply can’t because it’s a machine that can only measure movements, correlate and analyse them. It can’t observe or uncover the actual cause that makes the movement happen and it’s this (causation) that only small data can input on.
An example Lindstrom cited is from a certain American bank which noticed their customers were leaving in increasing numbers. Big data crunched the numbers and concluded via correlation that they were unhappy with the service they were receiving and that’s a fair first thought.
However it was small data (derived through humans simply asking the question why), which taps into the clues of why movement happens and uncovered the reason that they were simply going through a divorce, and that’s why they were leaving.
With only big data on hand, the marketing department may have crafted anything from simply writing to customers, looking for feedback on the service to making big, strategic changes to improve a service that may not have even been broken.
With small data in play and a knowledge of the cause, it can be as simple as offering understanding and presenting products that keep them as a customer.
Lack of creativity is the result of lack of boredom
There is a lack of creativity and innovation in the world and while it could be argued that is partly a result of completely new concepts being difficult to come across, Lindstrom argues that it’s more the outcome of a lack of boredom and the smart phone is the main evil culprit.
Hard to disagree.
See, there’s just no quiet time any more. I’m guilty and if you say you aren’t then you’re probably lying (sorry). You wake up, you reach for your phone. You’re at lunch, you look at your phone. You’re driving, you’re talking (not texting, right?) on your phone. You’re even on your phone watching TV or are at the computer screen. Hell, many people even combine the ultimate quiet time (bathroom) with a smart phone!
The times where you’d be silent, bored even, are evaporating and that means genuine, meditative think time, creation time, innovation time is evaporating along with it.
We’re increasingly less present with the world and what’s happening around us, unless it’s filtered through screens and that’s hurting intuition, innovation, engagement and stifling creative outcomes.
The application of these themes for all marketers should be at the top of strategy development.
Critical thinking should be applied to any new tech, innovation or invented theory because the danger in racing to that concept, while tempting, is that you ignore the balance required to make the whole strategy work
Observe, go out, and talk to the actual buyers of your proposition. The Japanese have whole concept based on this (Genchi Genbutsu) and they’ve used it to dominate production processes for decades. Don’t desk your plan without other inputs because it never works
Never listen to anyone who only presents a spreadsheet of data collected through a survey, lists or other means (unless it’s an outcome of the combination of the previous point).
These things have been and, I believe always will be, fundamental marketing frameworks.