Faced with overwhelming amounts of data, organizations across the world are looking at leveraging data and analytics (D&A) to derive insights to increase revenue, reduce costs, and mitigate risks. 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 . However, not many organizations are successful in transforming their data into insights despite being data-rich and having high D&A ambitions. In January of 2019, research advisory firm Gartner reported that 80% of D&A projects did not deliver business outcomes . While there are many reasons for this poor success rate, one key factor is that many firms struggle to effectively consume the insights derived from D&A.
But what exactly is an insight? Insight is the unknown elements such as relationships, patterns, categorization, inferences, prediction, and so on, if known will influence decision-making. These insights are typically derived using a combination of descriptive analytics, predictive analytics, and prescriptive analytics techniques. Descriptive analytics – “what happened” – analyzes historical data to identify past or lagging patterns. Predictive analytics – “what will happen” – forecasts future trends and events from historical data. Finally, Prescriptive analytics – “what will make it happen” – recommends the best course of action by using the insights derived from predictive analytics.
While there are many ways to classify an insight, from the D&A perspective, there are two types of insights: performance insights and actionable insights . Performance insights provide new visibility or knowledge of the measurement entity. 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, which are based on performance insights, are the insights that can be turned into action or response. An insight can be called actionable if it has three main characteristics.
- Actionable insights drive decisions.
- Actionable insights consume business resources like money, labor, and equipment to implement the decision.
- Actionable insights bring change in the business process when the decision is implemented.
With this backdrop, how can enterprises effectively deploy insights for actionable business outcomes? The FAAR framework can be deployed to increase the odds of insight consumption in organizations. FAAR is an acronym that is based on four factors: function, analysis, atomicity, and roles. The following section explains these four factors in detail.
Function states that insight consumption is dependent on the needs of the business stakeholders. Overall, there are three types of functions where stakeholders need insights in the organization: monitor, analyze, and details.
- C-suite and senior management need insights to monitor business performance.
- Managers need insights to analyze.
- Analysts and other individual contributors need insights at a very detailed level.
The second element in the FAAR framework is the level of analysis (i.e., the size and scale of the insights needed to measure and improve business performance). The level of insights for analysis can be at three levels: high, medium, and low.
- High-level analysis is abstract in nature and deals with insights into the business value chain.
- Medium-level analysis is focused on insights into specific processes in the value chain like order to cash (OTC), procure to pay (P2P), record-to-report (R2R), and so on.
- Low-level analysis describes insights into like activities and entities within the system.
The atomicity of insights consumed is the third element in the FAAR framework. From an atomicity perspective, insights can be of three types: granular, aggregate, and KPIs.
- Granular insights are detailed data or the lowest level of data and insights in the system. It includes both transactional data (like business events or actions such as purchase orders, invoices, and sales orders) and master data (on business entities such as products, suppliers, assets, and customers).
- Aggregate data is data about business categories or reference data elements such as manufacturing plants, customer groups, product categories, and stores.
- KPI (Key Performance Indicator) is the quantifiable measure used to evaluate the success of a measurement entity in meeting its performance objectives. The KPI could be a base measure like revenue or cost or a composite measure which is a combination of two or more base measures. For example, revenue is a base measure and gross profit that is derived from two base measures, i.e., sales and COGS (Cost of Goods Sold) is a composite measure or composite KPI. The measurement entity could be the entire enterprise, a business function (such as finance), a product category, a team, and more.
The last component in the FAAR framework is the roles. Access to insights includes the ability to create, edit, view, or share an insight and this is dependent on the business role of the insight consumer. Role-based access control (RBAC) or role-based security enables data and insight access to authorized business users based on their roles and responsibilities in the enterprise. This protects sensitive data and insights from unauthorized access and ensures that the right people have the right insights to perform their jobs as per the SoD (Segregation of Duties) policies of the organization.
So, how can organizations use the FAAR framework to implement D&A solutions? Based on the above four factors, insight can be consumed by the decision-makers using three main tools or mechanisms: dashboards, BI (business intelligence) reports, and transactional reports.
A dashboard offers a visual snapshot of business performance based on KPIs. Application of the FAAR Framework means that the purpose of the dashboard:
- Is for executives and C-suite members
- Has a low level of analysis
- Is for monitoring business performance
- Is based on KPIs
A BI report provides a consolidated view of business performance based on business categories or aggregates. Application of the FAAR framework means that the BI report:
- Is for manager or mid-level management staff
- Has a medium level of analysis
- Is for analyzes business performance
- Is based on business categories or aggregates like manufacturing plants, chart of accounts, customer account groups, and more
An OLTP (Online Transactional Processing) report from systems such as ERP (enterprise resource planning), EMR (Electronic Medical Record), and so on provides the most detailed or granular view of business performance. Application of the FAAR framework means that the OLTP report is:
- For analysts who need to gather and interpret the detailed level of data
- For a low level of analysis at the line-item or activity level
- For measuring performance at a detailed level
- Based on granular data
The FAAR framework with the three key implementation or deployment options is shown below.
While many D&A projects do a great job on the technical aspects, unfortunately, many projects do not have a clear understanding of the functional and business aspects of insight consumption. Knowing the function, roles, level of analysis, and atomicity of data can help enterprises determine an effective solution for consuming insights. However, the success of the FAAR framework rests on a good objective statement, quality data including the right governance mechanisms, a strong D&A model, and user education or training. Although the FAAR framework can sometimes oversimplify complex situations for deploying insights, it can still serve as a great starting point for organizations and D&A leaders and managers to formulate and simplify their insight consumption and deployment strategy.