An End to Useless KPIs: Data Driven Strategy

Becoming a “data-driven organization” has been a popular goal among CIOs, CTOs, and CDOs, but there is little agreement on what that actually looks like. Different parts of a business often have varying goals and expected outcomes, making it difficult to determine whether a data-driven project has truly succeeded at the organizational level.

Key Performance Indicators (KPIs) are central to answering questions like “Did we do what we intended?” and “What should we do next?” KPIs are meant to provide actionable data to improve organizational operations. However, a recent MIT Sloan survey of over 3,200 global executives found that nearly 30% said their organization’s KPIs only somewhat, minimally, or not at all drive how they lead or manage people and processes. Over 40% said their KPIs are only moderately influential.

KPIs have evolved from simple transactional benchmarks to complex behavioral and performance measures. With the rise of machine learning and AI, KPIs can now offer granular insights into organizational health. However, just because something can be measured doesn’t mean it should be, or that it’s relevant to the organization’s health. Many KPIs are valuable within specific functional areas, but often aren’t tied to the broader strategic vision. This disconnect can derail progress and mislead decision-makers, making it essential to rethink how data is used.

Feeding the (Analytics) Beast

The typical cycle is: KPIs generate data, analytics interpret the data, decisions are made, and KPIs measure the impact. The quality of decisions depends on the quality of data. Organizations now generate more data than they can handle, so leadership must focus on identifying and analyzing the right, relevant data to guide the organization.

To improve decision-making, organizations use predictive analytics (to forecast results) and prescriptive analytics (to recommend actions). While forecasting and decision-making are not new, technology has made these processes faster and more comprehensive. AI and machine learning have improved the accuracy and nuance of predictive models, provided the inputs and queries are valid. Predictive analytics can show what has happened and suggest what should or could happen next.

Leadership must then decide if these predictions matter and what actions to take. Most business intelligence tools stop at predicting outcomes, which is why prescriptive analytics—laying out what to do with predictive results—is gaining interest. Prescriptive analytics establishes rules or paths to optimize processes toward strategic goals. While only about 11% of large and mid-size organizations currently use prescriptive analytics, Gartner forecasts this will grow to 37% by 2022. Gartner also predicts that by 2020, predictive and prescriptive analytics will attract 40% of new enterprise investment in business intelligence and analytics.

Despite these advances, predictive and prescriptive analytics are often used in silos rather than across the entire organization. Leaders must avoid letting isolated data drive decisions that don’t align with the strategic vision. Every decision should support the organization’s overarching goals.

If KPIs are set in isolation (e.g., “What’s good for this department?”), they are unlikely to help achieve organization-wide objectives. KPIs should be measurements, not targets. Siloed data and KPIs prevent correlation between different parts of the organization and hinder progress toward health, growth, and sustainability.

The challenge is to align “what do we know?” with “what should we do next?” Leadership must commit to gathering and analyzing only information relevant to strategic goals, coherently and across silos.

The Case for Orchestrated KPIs (OKPIs)

The analytics market is expanding rapidly, offering tools to analyze every aspect of an organization. However, orchestrating these tools can be more challenging than acquiring them. Analytics orchestration harmonizes the collection and analysis of disparate data across the enterprise.

Applying this orchestration concept to KPIs means creating an “orchestration layer” of KPIs (OKPIs): cross-functional and cross-departmental KPIs that sit above business unit KPIs to interpret and align their meaning. This provides a holistic understanding of their significance and value to business outcomes. It unifies leaders to act cohesively, rather than having departments operate at different paces or with different priorities. With this clarity, leaders can set KPIs for their teams that align with both individual and company goals.

OKPIs are not hypothetical. Startups and bootstrapped companies increasingly use OKPIs to stay agile and avoid waste from misaligned metrics. This leads to better understanding of performance, more accurate market data, and more realistic milestones. Orchestrated KPIs ultimately provide a deeper understanding of products and markets.

Think “KPIs by Design” for More Standardized and Stronger Data

The OKPI layer is not a superficial addition that keeps data ambiguous. Instead, KPIs should be embedded early in the development process, refined and standardized with full buy-in from executives and leaders. Early codification ensures that KPI variants share commonalities as analytic architectures are built. Otherwise, KPIs reflect only individual experiences and fail to align with orchestrated KPIs, resulting in less coordination and insight.

Proactively embedding KPIs into processes and architectures leads to more progress toward intended results. Achieving a KPI is only valuable if it advances strategic priorities. Too often, organizations focus on achieving KPIs rather than improving on strategic goals, leading to wasted effort and reduced employee engagement.

A data-driven culture is foundational. Setting clear goals and expected outcomes is the first step in developing a data and analytics strategy. Companies must share KPIs across the organization, aligning everyone along a continuum of individual, team, and company goals. OKPIs align subordinate KPIs, resulting in strategic and operational improvement, a stronger analytical basis for decision-making, and enhanced focus on what matters most.

---

1. 1."Forecast Snapshot: Prescriptive Analytics Software, Worldwide, 2019," Gartner, January 2019
2. 2."Combine Predictive and Prescriptive Techniques to Solve Business Problems," Gartner, October 2018