You’ll be familiar with the incessant demand for insights. What once seemed a niche pursuit is now a universal demand. So what happened?
For marketers, it was the shift from traditional (slow, scarce, expensive) research techniques to the ready availability of a vast corpus of (digital) behavioural data. Today we can reasonably argue that the purpose of data science is to extract insights from data.
From traditional research to data science
Yet its important to put these two worlds into context. Traditional research has long relied on advanced statistical techniques from clustering to regression analysis and econometrics. Data Science, standing on the shoulders of these approaches, has turbo-charged our capacity to handle vast volumes of data at speed, while enhancing our ability to leverage both structured and unstructured data, and to deliver data driven automation.
Thinking more about this shift we also see the demand for accelerated speed-to-insight (as a competitive weapon), and especially the appetite for real time insight. Personally, I well remember the passionate discussion of the latter phrase in my time as Insight Director. There was clearly a sense of existential angst. But why?
To caricature the traditional research perspective, insights are the result of a special alchemy; fusing survey (quantitative) and focus group (qualitative) data along with economic, demographic, psychological, cultural and sociological perspectives. To this way of thinking insights aren’t a commodity to be accessed on demand, they are the rare and precious output of an opaque and mysterious synthesis – one that only certain highly qualified ministers may perform.
As such, the very concept of (democratised) real time insights was anathema – exposing a chasm between analogue research and the burgeoning digital model that has yet to be fully resolved.
A critical aspect of the best traditional approaches, one that is only slowly being addressed today, is the inherent respect for the human individual. This mindset recognises the complexity and capriciousness of human behaviour and decision making and acknowledges the importance of multiple interactions, channels and time horizons.
By contrast the digital model has too often shifted such complexity out-of-sight, considering it beyond the purview of digital marketers as they focus on abstracted metrics.
It is important to recognise that in many respects these two approaches to insight are asking different questions. In the case of the traditional approach we are far more likely to be asking the WHY (did that happen), while in the digital model we are much more comfortable with limiting our enquiry to the WHAT (happened).
Finally by way of overview it is worth reflecting on the term insight itself. Where hindsight looks backwards and foresight looks forwards, insight looks inward to the true nature of things, literally ‘seeing into a situation’.
Ultimately we need to move toward a segmentation of insights, with a language to match. You may have already have noted that many digital insights are little more than observations – statements of what has happened. But maybe its best not to rock the boat too much – so lets call these tactical insights.
These insights are not to be downplayed. Everything we do to optimise a campaign in-flight can be included here, these are the near real time observations that enable rapid, tactical, responses to optimise performance and to achieve the agreed performance indicators.
BI solutions have long delivered this level of analysis, enhanced by visualisation to support decision making. The key step today is to facilitate the real time data flows that enable rapid, tactical, responses to dynamic circumstances. The granularity of that response will be determined by the quality and range of data available.
This level of insight should be distributed and easily accessible to all stakeholders – democratising analytics and breaking the grip of the priesthood!
Yet what most people have in mind when they consider insights are the big ideas that often lie behind new businesses or new business strategies. They are game changing, transformative perspectives on a product, service or market.
Many definitions of insight include reference to intuition, sudden revelation, or similar. The beauty of strong data management and effective BI solutions is that this potential is radically democratised. With more people having access to the data, we significantly increase the chances of someone having that ‘aha’ moment.
Thinking about the current crisis, many businesses have had to leverage strategic insights to refine or develop their proposition for the new paradigm. Obvious examples include the need to develop effective zero touch and remote customer experiences, but underpinning these responses the crisis also exposed the need to develop customer understanding – too many businesses were caught out as they simply did not know enough about their customer when the crisis first hit.
Strategic insights aren’t about optimising on the fly in the way of tactical insights, and while the need for speed and agility is a universal requirement today, the quality and depth of insight must be the primary driver.
Strategic Insights can be challenging and disruptive, you may need to work hard to embed them within an organisational culture. As such it is worth considering the six pillars of sticky ideas identified by Chip and Dan Heath in their book ‘Made to Stick’. They are:
Simple, Unexpected, Concrete, Credible, Emotional, Tells-a-story
Strategic decisions are difficult to undo, tactical decisions can be undone very quickly, this is the vital difference to have in mind. The cumulative effect of multiple tactical decisions can be enormous, but the effect of a single strategic decision is enormous in itself.
We’ve danced around an actual definition of insight – e.g. a true, accurate or deep understanding of someone/or something – but in our world there is one key dimension that you will all recognise.
All insights must be actionable – they must enable us to respond.
To return to our original question, are insights our raison d’etre?
Yes, they must be, and they must be the driving force behind our data investments. We can and should challenge our data teams to tell us new things, to present new associations and correlations and to challenge conventional thinking.
After all data driven decision making is a misnomer, data is a means to an end, the end is insight and the objective is improved performance.
With that in mind we can build new ways of integrating data into the core of our approach, with managers and leaders working with data scientists at a strategic level to test hypotheses, explore predictive models and deliver prescriptive solutions. In parallel our marketers can work with our BI solutions to provide tactical insights and actions.