Demystifying Actionable Insights in Data and Analytics

By on
Read more about author Prashanth H. Southekal.

Every company today is a data company. Organizations worldwide are striving to derive business insights from the enormous amounts of data that are captured and stored to measure and improve business performance. McKinsey found that insight-driven companies report EBITDA (earnings before interest, taxes, depreciation, and amortization) increases of up to 25%. According to Forrester, organizations that use data and insights for decision-making are almost three times more likely to achieve double-digit growth.

But what exactly is an insight? In simple words, insights are the unknown elements such as relationships, patterns, categorization, inferences, predictions, trends, outliers, and so on that – if known – will influence the decision. While there are many ways to classify an insight, from the data and analytics perspective there are two types of insights.

Performance insights: Performance insights provide new visibility or knowledge of the entity that is subject to measurement and performance. Examples of performance insights include the top three SKUs (Stock Keeping Units) by sales quantity, the top five customers by CLV (customer lifetime value), and so on. Performance insights can be generated by data scientists or even with generative AI tools like ChatGPT.

Actionable insights: Actionable insights, which are based on performance insights, are the insights that can be turned into action or response. An insight can be termed actionable if it has three characteristics.

  • Actionable insights drive decisions.
  • Actionable insights consume business resources like, money, labor, information, equipment, etc. to implement the decision.
  • Actionable insights bring change in the business process when the decision is implemented.

While performance insights provide new knowledge and are foundational, actionable insights matter most. So, how can enterprises maximize the value of these two types of insights – performance insights and actionable insights? In data and analytics, “last mile analytics” links the insights to real business outcomes. In simple words, to achieve business results, business enterprises must focus on converting performance insights into actionable insights. However, this is not simple and straightforward. Below are five key steps to achieve this.

Step 1: Derive performance insights from the KPIs.

Leverage the business KPIs (key performance indicators) and derive meaningful performance insights to get visibility into past, current, and future states. Generating performance insights is very complex and expensive, as it requires looking at the data from different lenses such as time periods, customer demographics, location, market conditions, and more. To ensure that the effort that is invested in deriving performance insights is useful, ask these six key questions.

  1. Why do you want to know? What are the value levers and value drivers to know these insights?
  2. How much do you want to know? Performance insights should be derived from the current data because decisions must be based on the insights that are relevant in the current context.
  3. What is the value of knowing and not knowing?
  4. Who owns this KPI? Can you realize the change?
  5. Do you have the quality data to calculate the KPI?
  6. Have you addressed the framing bias by reframing the problem in at least three different ways encompassing different stakeholder perspectives, time frames, and locations?

Step 2: Formulate the decision problem.

At the highest level, performance “nice to know” insights are actionable when tied to three main strategic business decisions: increasing revenue, reducing cost, and mitigating risk. In this regard, a typical decision problem has four key elements: objectives, alternatives, outcomes, and payoff.

  1. The objective defines measurable actions the business intends to achieve.
  2. A decision based on performance insights exists only when there are alternatives (i.e., potential options are based on different performance criteria such as profit margin, cost, time, quality, service, and more).
  3. The outcomes are the resulting situations that would arise by pursuing the selected alternatives.
  4. The payoffs are the benefits placed on the outcomes associated with each alternative.

The Pugh Matrix or the DEAR Model can help the business decide on the best alternative.

Step 3: Develop an eye for risk.

Almost any decision worth doing is inherently risky. As Mark Zuckerberg, CEO of Facebook once said, “The biggest risk is not taking any risk.” Even if the decision is well-thought-out using the Pugh Matrix or the DEAR Model, it will still have some risks. Overall, there are five main types of risk a business can face from implementing the alternative:

  • Strategic risk
  • Compliance risk
  • Operational risk
  • Financial risk
  • Reputational risk

To mitigate the risk, collect inputs from diverse and multiple experts from various lines of business on where the decision could go wrong if the alternative or solution is implemented. Put on your pessimist’s hat or imagine worst-case scenarios for a while and make a list of everything that could go wrong. Also, working with multiple, diverse experts will help avoid the confirmation bias (i.e., cherry-picking information that confirms the existing beliefs or hypothesis). 

Once the risks are identified, the scope of the risks needs to be thoroughly analyzed from both positive and negative perspectives based on severity (S), occurrence (O), and detection (D).

  • Severity is the potential effect of the failure on a scale of 1 (minimal impact) to 5 (high impact)
  • Occurrence rates the likelihood that the failure or loss will occur from the risk on a scale of 1 (very unlikely) to 5 (very likely).
  • Detection rates the likelihood that the problem will be detected before it reaches the end-user/customer on a scale of 1 (low chance) to 5 (high chance).

Now, create a risk scorecard by multiplying the three numbers to get an overall risk score called Risk Priority Number (RPN), which can then be used to rank and prioritize the risks. For example, for the risk item “CFO Tracy could leave the company”, if the severity score is 3, occurrence score is 4, the and the detection score is 5, then the RPN score is 4 x 3 x 5 = 60. The risk scorecard will give you a list of all identified risks that are prioritized for mitigation and resolution.

Customer XYZ could pay the invoice late 25330
CFO Tracy could leave the company34560
Competitor will undercut the price52110

When it comes to risks, many enterprises tend to concentrate on things that can go wrong and run into analysis-paralysis mode and fail to capitalize on the opportunities. But sometimes it is even good to take a risk, as it pushes your business to go outside of its comfort zone and become stronger and better. Basically, some risks are good and desired, while some need to be eliminated or contained as much as possible. The strategy for risk management or mitigation is to address the risk in one of four ways.

  1. Avoidance
  2. Retention
  3. Transferring or sharing
  4. Reduction or loss prevention

Step 4: Identify the necessary resources to execute the decision.

Once the best alternative to executing the decision is selected and validated with risk mitigation strategies, then pertinent business resources should be mobilized to execute the decision. Resources in business are the basic building blocks of the organization to achieve its objective. There are four main types:

  1. Financial resources 
  2. Human resources 
  3. Material resources
  4. Intellectual resources. 

Resource examples include tangible assets, such as its plant, equipment, finances, and people, and intangible assets, such as technology, data, patents, copyrights, brands, and reputation. Overall, the selection and allocation of the resource to make the insights actionable depends on the scale and significance of the decision.

Step 5: Manage change.

Deploying the resources to implement the decision or the actionable insight is a complex process and involves changes to the business model of the organization. Managing change requires strong leadership, effective communication, and validation with appropriate feedback mechanisms or governance so that the right people are managing the change pertaining to the insights in the right manner.

“Actionable insights” is not a buzzword. At their core, actionable insights are about leveraging data to measure and improve business performance. In today’s data-centric economy, data and analytics can transform businesses by providing insights for sound decision-making and improved business results. While most businesses spend a lot of time and effort in capturing and storing data, what is more important is using the data collected to derive actionable insights and improve the performance of the business – to increase revenue, reduce expenses, and mitigate risk.