Welcome to the beginning of a journey that, we hope, will reshape how you think about, measure, analyze and influence business performance. If you are more of a visual learner, we encourage you to start by watching Abhi's talk at Data Council 2023 called Designing and Building Metric Trees. From there, you'll be ready to get into the thick of it.
Every organization is fundamentally a system of inputs and outputs.
They exist to convert capital—physical, human, and intellectual—into some kind of value, most commonly cash. This transformation process is central to how organizations operate and succeed. It is the fingerprint of a business.
It’s no coincidence that most successful businesses embrace this perspective and cultivate a common understanding of the system from the leadership down to the individual levels. These businesses succeed because every member in the organization shares an understanding of their contributions towards achieving a larger, overarching goal, often referred to as a "North Star." This collective understanding is what defines a growth model within a company.
A growth model is a conceptual blueprint to help businesses understand, reason about and communicate the mechanisms and strategies for driving growth. It is essentially a causal model that allows members to anticipate results based on certain actions and, conversely, deduce actions from the results obtained. The model reduces uncertainty from strategic planning and decision-making and makes operational adjustments easier to perform and track.
There are many ways to express a growth model. Most are visual, with abstract representations of inflows and outflows. Others are concrete mathematical models that hide their complexity in spreadsheet formulas. While visual diagrams are helpful in high-level communication, and spreadsheet models offer the perception of finite precision, their intangibility is limiting for most stakeholders. This makes integrating them into day-to-day operations is hard, if not impossible.
The purpose of this guide is to help you translate the abstract elements of your growth model into a concrete, measurable, and interactive format. Put simply, we want to help you reify your growth model in the form of a metric tree. You'll learn how a metric tree can serve as a practical bridge, connecting your overarching growth strategy to daily tasks such as causal analysis, forecasting, planning, and business reviews.
Introducing Metric Trees
Metric trees are surprisingly powerful despite their apparent simplicity. If you’ve ever tried to connect the conceptual “physics” of a business to a concrete action, you know it's easy to get lost on the ladder of abstraction.
Metric trees bridge this gap thanks to a few useful characteristics, which are summarized in their name.
Metrics are abstract representations of business events, either actual or hypothetical. As such, they exist in the context in which the events take place. Each metric embodies a “cube” of dimensions including time as well as those related to the entities involved in an event. This means they are great for presenting the “facts” about a business without getting bogged down in the minutia of the underlying data.
On the flip side, metrics are limited by their abstraction. They only represent facts to the extent that people can understand and agree upon their definitions. If there is ambiguity or disagreement about a metric’s semantics, its utility is compromised. This is particularly true when definitions differ across teams or business units.
In essence, metrics are only as useful as the clarity of their definitions and consensus about their importance.
In essence, metrics are only as useful as the clarity of their definitions and consensus about their importance. This is where trees come in.
A tree is the structural backbone that binds every metric to its business significance. Metric trees are constructed “backward,” starting with the ultimate output (the North Star) down to the most distant input. Each layer of metrics is derived from decomposing those above it. This means there is one, and only one, definition for each metric, and their significance is inherited from the generations above. To put it plainly, every metric derives legitimacy from its lineage back to the North Star.
As the metrics on the tree diverge, the scope of their significance narrows—functions, then teams, then processes, and individuals—but their connection to the greater system is never lost. At the terminus of every branch is the interface between analysis and action. From that point, the tree provides a causal chain all the way back to each action’s impact on the North Star.
So, what is a metric tree?
A metric tree is an analytical model that takes metrics and their weighted relationships to represent the flow of inputs into a business and how they are transformed into outputs.
A metric tree is an analytical model that takes metrics and their weighted relationships to represent the flow of inputs (physical capital, human capital, intellectual capital, etc.) into a business and how they are transformed into outputs (including financial, customer, and strategic value outputs).
Put simply, a metric tree is the logical representation of a business’s growth model.
What does a metric tree look like?
Metric trees use diagrams to summarize how inputs contribute to outputs (and ideally, how inputs lead to the North Star Metric).
At the top, you find outputs, and it all starts with $$$’s – with revenue. These outputs decompose into inputs. Inputs consist of components that are mathematical identities (e.g. Leads and Win Rate) and influences (Lead score).
Any input that is also an influence metric has a contingent relationship, both with components below and with outputs above. These relationships have properties, for example, “Strength” or “Confidence”.
Finally, inputs have dimensions, for example, Market Segment or Country.
Metric trees can also be nested, with layers of analysis as outputs become inputs that feed deeper into a larger, more complex profit model.
Applications and Use Cases
Using a metric tree to describe a business as a function enables us to take advantage of the inherent precision and predictability of mathematical analysis to help optimize business decision-making – providing grounding for informed decisions.
Metric trees also mitigate the problem of information overload. Whether it’s intra-departmental or at the leadership level, it’s common that data interpretation becomes somewhat haphazard and disconnected from the bottom line, creating a void between inputs and measuring success.
In contrast, metric trees provide a framework that allows leaders to decompose complex business functions into smaller sections that are more comprehensive.
This conceptual structure becomes a central reference point to understand and evaluate the relationships within a business. That said, metric trees do not dumb down the analysis: with them, it is as easy to describe the entire business as a single unit or function.
A balance of simplicity and comprehensiveness makes it an excellent tool for decision-making at a holistic level. A few simple use cases include:
· Root cause analysis: A metric tree can help guide the analytical process to ensure that companies can understand what really impacts any given output – for example, are win rates the real cause for low revenue, or is it the ASP?
· Employee and stakeholder onboarding: Metric trees represent the fingerprint of an organization and are a powerful way to explain to newcomers how a company’s growth model works.
· Mission statement for data teams: Metric trees serve as a navigational tool that helps data teams establish the required metrics while providing a path to finding a company’s growth levers – and how to pull these levers.
· Dashboarding: a fleshed-out metric tree can help companies build dashboards, grouping metrics into related areas (e.g., revenue vs. leads, customers, win rate, ASP, etc).
· Faster business reviews and forecasting: With all necessary metrics, influences and levers available at a glance, companies can go through the business review process more accurately and faster.
Steps to constructing a metric tree
Metric trees are, at first glance, a rather simple concept. It’s also true that metric trees are easy to use. So yes, metric trees appear simple because it’s necessary to have a tool that stakeholders can easily interpret for action. However, metric trees reflect underlying business processes and relationships that are not simple.
Just like modeling any business process, the devil is in the details. Taming these details requires a regimented, step-by-step process, and the decisions you make along this process dictate the utility of the metric tree – and its impact.
To capture, visualize, and understand the key metrics that drive the success of a company or unit we suggest that you follow three essential steps:
· Defining the North Star metric: The North Star metric is the primary measure of success for the company or unit. This metric should align with the long-term value creation for the company. This metric provides direction and focus for the entire organization.
· Decomposing components: The next step involves breaking down the North Star metric into its component parts, identifying the input metrics that directly contribute to the calculation of the North Star Metric and other key output metrics.
· Identifying influences: The final step involves identifying influence metrics, the metrics that have correlative, causal, or qualitative relationship with output metrics – in order to understand how changes in certain metrics can influence the performance of other metrics.
The goal of constructing a metric tree is to provide a clear, actionable framework that guides decision-making and growth strategies by highlighting the relationships between various metrics and identifying levers for influencing key outcomes.
How (and where) to design a metric tree
You can design a metric tree with any diagramming tool you find easy to use. In short, tools like Miro, Figma, or even MS Paint can be used for creating metric trees. The key is focusing on the outcome, not the design tool. Metric trees help visualize, communicate, and align a business's value creation and flow, enabling quicker issue resolution, accurate forecasting, and integrated planning.
The ideal metric tree setup for your organization is one that effectively supports understanding and managing value flow.
Ultimately, the effectiveness of metric trees extends beyond visualization to being integrated into data and models. Technically, these "trees" are more accurately described as graphs.
