Impara Building Metric Trees in Python

Metric trees are a powerful way to break down complex business goals into actionable, measurable components. When you are faced with a high-level objective—such as increasing revenue or improving user engagement—a metric tree helps you visualize how this goal can be decomposed into smaller, related metrics. Each branch of the tree represents a relationship: parent metrics are influenced by their children, and understanding these dependencies allows you to identify which areas to focus on for improvement. By systematically mapping out metrics in a hierarchical structure, you make it easier to assign responsibility, track progress, and ensure alignment across teams.


              123456789101112131415161718192021222324
            
import pandas as pd

# Define the structure of a metric tree for a business goal: Increase Revenue
data = {
    "Metric": [
        "Revenue",
        "Number of Customers",
        "Average Revenue per Customer",
        "Number of Website Visitors",
        "Conversion Rate",
        "Average Order Value"
    ],
    "Parent": [
        None,              # Revenue is the root
        "Revenue",         # Number of Customers drives Revenue
        "Revenue",         # ARPU drives Revenue
        "Number of Customers", # Visitors drive Customers
        "Number of Customers", # Conversion Rate drives Customers
        "Average Revenue per Customer" # AOV drives ARPU
    ]
}

metric_tree_df = pd.DataFrame(data)
print(metric_tree_df)


              12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061
            
import matplotlib.pyplot as plt
import matplotlib.patches as patches
import pandas as pd

# Manual coordinates for clean diagram layout
node_positions = {
    "Revenue": (0, 0),
    "Number of Customers": (-3, -2),
    "Average Revenue per Customer": (3, -2),
    "Number of Website Visitors": (-5, -4),
    "Conversion Rate": (-1, -4),
    "Average Order Value": (3, -4)
}

edges = [
    ("Revenue", "Number of Customers"),
    ("Revenue", "Average Revenue per Customer"),
    ("Number of Customers", "Number of Website Visitors"),
    ("Number of Customers", "Conversion Rate"),
    ("Average Revenue per Customer", "Average Order Value"),
]

def plot_metric_tree_clean(node_positions, edges):
    fig, ax = plt.subplots(figsize=(12, 7))

    # Draw edges (behind everything)
    for parent, child in edges:
        px, py = node_positions[parent]
        cx, cy = node_positions[child]
        ax.plot(
            [px, cx], [py - 0.3, cy + 0.3],
            color="gray", linewidth=2, zorder=1
        )

    # Draw nodes (in front)
    for metric, (x, y) in node_positions.items():
        rect = patches.FancyBboxPatch(
            (x - 1.35, y - 0.125),
            2.7, 0.25,
            boxstyle="round,pad=0.3",
            linewidth=1.5,
            edgecolor="#1976D2",
            facecolor="#1976D2",
            zorder=3
        )
        ax.add_patch(rect)

        ax.text(
            x, y,
            metric,
            ha="center", va="center",
            color="white", fontsize=11, fontweight="bold",
            zorder=4
        )

    ax.axis("off")
    plt.title("Metric Tree: Revenue Breakdown", fontsize=16)
    plt.tight_layout()
    plt.show()

plot_metric_tree_clean(node_positions, edges)

Metric trees are especially valuable for diagnosing business performance issues. When a core metric—such as revenue—falls short of expectations, the tree structure guides you through the hierarchy to pinpoint where the problem might originate. For instance, if revenue is underperforming, you can examine its immediate drivers: is the number of customers declining, or is the average revenue per customer dropping? By following each branch down to more granular metrics, you can quickly isolate which component is responsible for the change. This structured approach not only accelerates root cause analysis but also helps you design targeted interventions, ensuring that your efforts are focused where they will have the greatest impact.

Tutto è chiaro?

Grazie per i tuoi commenti!

Sezione 6. Capitolo 2

Chieda ad AI

Chieda pure quello che desidera o provi una delle domande suggerite per iniziare la nostra conversazione

Suggested prompts:

Can you explain how to build a metric tree for a different business goal?

How do I interpret the relationships between parent and child metrics in the tree?

Can you give an example of how to use this metric tree to diagnose a business issue?

Scorri per mostrare il menu


              123456789101112131415161718192021222324
            
import pandas as pd

# Define the structure of a metric tree for a business goal: Increase Revenue
data = {
    "Metric": [
        "Revenue",
        "Number of Customers",
        "Average Revenue per Customer",
        "Number of Website Visitors",
        "Conversion Rate",
        "Average Order Value"
    ],
    "Parent": [
        None,              # Revenue is the root
        "Revenue",         # Number of Customers drives Revenue
        "Revenue",         # ARPU drives Revenue
        "Number of Customers", # Visitors drive Customers
        "Number of Customers", # Conversion Rate drives Customers
        "Average Revenue per Customer" # AOV drives ARPU
    ]
}

metric_tree_df = pd.DataFrame(data)
print(metric_tree_df)


              12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061
            
import matplotlib.pyplot as plt
import matplotlib.patches as patches
import pandas as pd

# Manual coordinates for clean diagram layout
node_positions = {
    "Revenue": (0, 0),
    "Number of Customers": (-3, -2),
    "Average Revenue per Customer": (3, -2),
    "Number of Website Visitors": (-5, -4),
    "Conversion Rate": (-1, -4),
    "Average Order Value": (3, -4)
}

edges = [
    ("Revenue", "Number of Customers"),
    ("Revenue", "Average Revenue per Customer"),
    ("Number of Customers", "Number of Website Visitors"),
    ("Number of Customers", "Conversion Rate"),
    ("Average Revenue per Customer", "Average Order Value"),
]

def plot_metric_tree_clean(node_positions, edges):
    fig, ax = plt.subplots(figsize=(12, 7))

    # Draw edges (behind everything)
    for parent, child in edges:
        px, py = node_positions[parent]
        cx, cy = node_positions[child]
        ax.plot(
            [px, cx], [py - 0.3, cy + 0.3],
            color="gray", linewidth=2, zorder=1
        )

    # Draw nodes (in front)
    for metric, (x, y) in node_positions.items():
        rect = patches.FancyBboxPatch(
            (x - 1.35, y - 0.125),
            2.7, 0.25,
            boxstyle="round,pad=0.3",
            linewidth=1.5,
            edgecolor="#1976D2",
            facecolor="#1976D2",
            zorder=3
        )
        ax.add_patch(rect)

        ax.text(
            x, y,
            metric,
            ha="center", va="center",
            color="white", fontsize=11, fontweight="bold",
            zorder=4
        )

    ax.axis("off")
    plt.title("Metric Tree: Revenue Breakdown", fontsize=16)
    plt.tight_layout()
    plt.show()

plot_metric_tree_clean(node_positions, edges)

Tutto è chiaro?

Grazie per i tuoi commenti!

Sezione 6. Capitolo 2