Data, data everywhere, but not any time to think

Despite the plentiful, almost overwhelming, amount of data available to companies today, it can be challenging to translate this wealth of information into initiatives that successfully improve business performance.

Imagine a pipe that delivers your various data sources, e.g., EPOS data, panel data, rich data, market intelligence, commercial performance, inventory holding. The pressure in this pipe has been growing constantly over the last 10 years as we are receiving more data, and new types of data, more quickly than ever before. Unless the internal system has been adjusted to receive and manage this influx it’s possible to drown, rather than flourish.

Automating data management can help manage the flow, but often a fundamental reengineering of the decision-making process – a core element of any operating model – is required to capitalise on the enriched insights that can be mined. Indeed, a zero-based approach to decision-making can help define the process while reducing the overall cost of data, which for many consumer products companies and retailers is not a trivial sum.

My first decision-making experiences were in the military. In that environment it was critical to operate within a faster decision-making cycle than your adversary. The emphasis was on collecting information, creating intelligence, building a plan to 80% and then executing it – and quickly adapting it as the situation unfolded. The data collection plan was informed by our strategic intent and short -term tactical situation. The right information was passed through the organisation to the planners who needed it, who in turn fed their intelligence into the well-oiled planning and execution machine.

Applying this approach to a commercial setting highlights two simple but critical questions:

1. What decisions do I need to make?
2. How do I make them?

What decisions do I need to make?

Assuming they align to the corporate strategy, BU and Functional targets and KPIs are a good starting point, e.g., sales from new products, marketing ROI, reduced inventory holdings. The levers by which you deliver these KPIs, e.g., NPD, pricing, demand and supply planning are the processes within which you establish the key decisions points. The data requirement, which should be quite specific, is informed by the nature and frequency of those decision points.

This works well for proactive initiatives. However, reacting to broader market dynamics relies on monitoring the activity of others and maintaining a general level of awareness. This can be a data-hungry process so it is important to determine what will help you identify changes that are remarkable and impactful.

Ultimately, the data you are paying to receive, analyse and distribute should be informing performance improvement. If it isn’t, or it is a secondary or tertiary support tool, you should consider re-purposing the associated funding.

How do I make them?

We can break down this question further into Who?: person or machine? And How?: via what mechanism?

Many decisions can and should be automated to efficiently deliver a range of effects, such as customer experience (rapid credit checking) or dynamic pricing (airline ticketing). While AI is taking strides forward, e.g. Mars Wrigley and Google’s Maltesers AI cake or Unilever’s harnessing of unstructured data to inform marketing choices, humans, leveraging experience, remain the default decision makers and logic programmers.

Defining how decisions are made is more than the completion of a RACI model and establishing a stage gate process. It is an output of the prevailing culture and specifically the level of empowerment that exists within the organisation. Reducing bureaucracy and red tape, breaking down silos, and enforcing performance management / accountability for decisions taken are key elements of rapid, empowered decision making.

Performance improvement and potential competitive advantage is available to those organisations that can collect and act on data and insight ahead of their competition. To do this well requires an understanding of the critical decisions, a rich, targeted data set, a deliberately engineered process, and a culture of empowerment. And perhaps a little good fortune.