“A lot of times data teams focus too much on things that are not influenceable. There are a lot of dashboards that only show KPIs that are not directly influenceable. That’s not good because they are not actionable.”
Every data initiative starts with the best intentions: track everything, measure progress, drive decisions. Yet too often, we end up with a maze of metrics that confuse rather than clarify.
What if there was a structured way to cut through the complexity? A framework that not only organizes your metrics but also reveals their true strategic value?
This five-layer framework for metric classification will transform how your organization approaches KPIs – from selection to implementation to action.
Why Metric Classification Matters
Building a data-driven organization is a lot about understanding which metrics matter and why.
Proper metric classification helps you:
- Focus on truly actionable metrics instead of vanity metrics
- Identify gaps in your data strategy
- Align your metrics with your company’s current stage and objectives
- Prevent over-investment in tracking deprecated metrics
- Create a shared language for discussing KPIs across teams
Let’s dive into five powerful ways to classify your metrics.
1. Influenceability: Can You Actually Move The Needle?
The most fundamental classification splits metrics into two categories:
→ Directly influenceable metrics
For example:
- Ad impressions
- Discount rates
- Email campaign frequency
→ Non-directly influenceable metrics
- Overall profit
- Total revenue
- Market share
- Brand awareness
The key insight? You need both types, but they serve different purposes.
Influenceable metrics are your levers for change, while non-influenceable metrics are your ultimate goals.
The art lies in modeling the relationship between them (for example with a KPI Tree).
2. The Pirate Framework: AARRR for Different Company Stages
This classic startup framework breaks down metrics into five categories that sound like a pirate’s favorite word (AARRR): 🏴☠️
- Acquisition: Getting users into your funnel
- Activation: Converting users to key behaviors
- Revenue: Turning activity into money
- Retention: Keeping and growing customers
- Referral: Driving viral growth
Your company’s stage determines which metrics deserve the most attention. Early-stage startups usually focus heavily on acquisition and activation, while more mature companies typically shift focus toward retention and profitability.
It usually doesn’t make much sense for your data team to focus on profitability metrics, if your company is in hyper growth stage.
3. Time Views: When Do You Measure?
Understanding how metrics are attributed to different dates is crucial.
There are four distinct approaches:
→ Transactional
- Measured at event date
- Never changes retrospectively
- Perfect for day-to-day operations
- Subject to seasonality
Example: New Customers reported at customer creation date.
→ Cohorted
- Measured at a past related event
- Can change retrospectively
- Ideal for understanding business improvement
- More complex to understand
Example: Cumulated Customer Lifetime Value reported at the date of a customer’s first order.
→ Portfolio
- Measured at period start/end
- Great for asset tracking
- Perfect for external reporting
- Less useful for quick decisions
Example: Number of Customers at the end of each month.
→ Forward Attributed
- Measured at future related event
- Ideal for marketing attribution
- Complex to understand
- Great for (marketing) budget allocation
Example: Cost per Order reported at the date of the order (even though costs were incurred before that, at the time of the click on an ad).
4. Prioritization: Not All Metrics Are Created Equal
Create clarity by assigning clear roles to your metrics:
- ⭐️ North Star KPI: Measures value delivered to market (e.g., nights booked for Airbnb)
- 🔝Top KPIs: Measures value captured from market (e.g., revenue). This is often confused with North Star KPIs. Remember: North Star KPIs describe the value a customer receives from the company while Top KPIs describe the value the company receives from customers.
- ✅ Focus KPIs: Currently being actively improved
- 🩺 Health KPIs: Monitored but not actively optimized. Think of it as tyre pressure or blood pressure. You won’t actively optimize them but you want to keep an eye on them and turn them into focus KPIs if the move beyond a certain threshold.
- 1️⃣ One Metric That Matters: The single most important influenceable metric right now. During the pandemic, this was the number of full homes listings in non-urban areas for Airbnb as demand shifted from holiday homes to work-from-home places.
- 😴 Sleeper KPIs: Currently unimportant but may matter later
- 👻 Deprecated KPIs: No longer relevant. A large number of these indicates that you may have built too much too early as your company pivots.
5. Metric Types: Raw vs. Composite
The final classification is straightforward but crucial:
→ Raw Metrics
- Not calculated from other metrics
- Pure measurement data
- Example: Number of orders
→ Composite Metrics
- Calculated from other metrics
- Derived measurements
- Example: Conversion rate (orders/sessions)
Bottom Line: Turn Classification Into Action
Stop drowning in a sea of undifferentiated metrics. Start by classifying your current KPIs using these five frameworks. You’ll quickly identify:
- Which metrics you can actually influence
- Where you might be focusing on the wrong stage of your customer journey
- How to structure your reporting for maximum impact
- Which metrics you can safely retire
Remember: The goal isn’t to track everything – it’s to track the right things in the right way. Your next dashboard might have fewer metrics, but they’ll be the ones that actually drive decisions.
Take action now: Pick your top five metrics and classify them using each framework. You might be surprised by what you learn.
