Start with a Metrics Model

Start with a Metrics Model'

Dinker Charak

Aug 04, 2023

The modern world of business is data-driven, with metrics serving as crucial signposts on the journey towards enhanced outcomes. However, the effectiveness of these metrics often hinges on their careful modeling, which should be tailored to reflect the unique dynamics of an organisation and its industry. A well-developed model can illuminate paths to success, provide warning signs of impending failure, and offer guidance for remediation.

Among the numerous metrics frameworks available, the Engineering Excellence to Business Outcomes (EEBO) Metrics framework stands out for its comprehensive approach to modeling a metric.

Characteristics of a Metrics Model

When applying the EEBO Metrics framework, there are several key characteristics that drive the Modeling:

  1. Domain-Based Reference: The reference that acts as a guidance for any metric should stem from the domain of the product and the industry of the organisation. This ensures the metrics goals are contextually relevant and achievable.
  2. Clear Success Criteria: The model should make success criteria transparent, enabling all stakeholders to understand when objectives need to be met.
  3. Failure Threshold: Similarly, it's vital to articulate a clear failure threshold beyond which either a pivot or termination is necessary. This helps avoid prolonged engagement with underperforming initiatives.
  4. Remediation Plans: Well modeled metric models should include a structured remediation plan for course correction when underperformance is identified. This avoids knee-jerk reactions and promotes thoughtful, strategic action.

Building a Comprehensive Metrics Model

Often, teams limit their focus to two fundamental components of a metric: face value and unit, such as "Our Lead Time is 5 days." However, this approach leads to an incomplete model, limiting the potential benefits of metrics usage. To fully leverage the power of metrics, organizations should consider the following components:

  1. Success Criteria: Define the goal to be achieved by using the metric. This sets a target and enables meaningful performance assessment.
  2. Domain Reference: Identify the typical value among similar products or services in your domain. This provides a benchmark for comparison and can spotlight opportunities for competitive differentiation.
  3. Baseline: Determine the team's current position based on an agreed-upon duration. This helps quantify subsequent progress and strides towards improvement.
  4. Failure Threshold: Decide on a value, persisting for an agreed-upon duration, at which pivots or sunk cost considerations will come into play. This serves as an early warning system, preventing prolonged engagement with problematic initiatives.
  5. Remediation Plan: Establish remediation guidelines to be considered during the modeling process. These guidelines can help navigate crisis situations without resorting to panic-driven decisions.
  6. Cost of Metric: The cost of maintaining a metric should be estimated based on the required infrastructure, complexity of code, and potential manual work or interventions. This can help organizations avoid costly overheads of tracking irrelevant or defunct metrics.


In conclusion, when deploying an EEBO Metrics framework, organizations should not only consider face values and units but also take into account success criteria, domain references, baselines, failure thresholds, remediation plans, and cost considerations.

This comprehensive approach to metric modeling will not only provide a deeper understanding of performance but also pave the way for meaningful and sustained improvements in business outcomes.