Navigator
Startup
Analytics Fundamentals

Metrics vs. Dimensions: What You're Actually Measuring

Understanding the difference between metrics (numbers that change) and dimensions (categories that don't) is the foundation of all analytics.

• Navigator Team
metrics dimensions data structure analytics basics

You walk into a meeting and someone says: “Our conversion rate is 2.5%.”

You nod. You know what that means. But do you?

Let’s break it down.

“Conversion rate” is a metric. It’s a number that changes based on conditions.

“2.5%” is the value of that metric right now.

But a conversion rate is actually made up of two things:

  • Numerator: 100 conversions (the metric: count of conversions)
  • Denominator: 4,000 visitors (another metric: count of visitors)
  • Result: 2.5% (a calculated metric)

Now, if you want to understand why your conversion rate is 2.5%, you need to slice it by dimensions.

What is a dimension? It’s a categorical attribute that doesn’t change. It’s a label.

Examples of dimensions:

  • Traffic source: Organic, Paid, Direct, Referral
  • Device: Mobile, Desktop, Tablet
  • Geography: US, Canada, UK, Germany
  • Product: Growth Plan, Pro Plan, Enterprise Plan
  • Campaign: Q4_PROMO_2024, Black_Friday, Launch_Sale

When you combine a metric with a dimension, you get actionable insight.

Metrics: Numbers That Move

A metric is any value that can be counted, summed, or calculated.

Count Metrics (How many?):

  • Number of visitors
  • Number of conversions
  • Number of signups
  • Number of support tickets

Sum Metrics (How much total?):

  • Total revenue
  • Total ad spend
  • Total time spent on site (summed across all users)

Average Metrics (What’s typical?):

  • Average order value (total revenue / number of orders)
  • Average session length
  • Average customer lifetime value

Rate Metrics (What’s the relationship?):

  • Conversion rate (conversions / visitors)
  • Churn rate (customers lost / total customers)
  • Click-through rate (clicks / impressions)

Ratio Metrics (How do two things compare?):

  • CAC payback period (months of revenue / CAC)
  • LTV:CAC ratio (lifetime value / customer acquisition cost)
  • Return on ad spend (revenue / ad spend)

The key thing about metrics: They vary. They go up and down. Last week your conversion rate was 2.3%. This week it’s 2.5%. That variation is what makes them useful—you can see trends and changes.

Dimensions: Labels That Organize

A dimension is a categorical attribute. It doesn’t “vary” in the way a metric does. It’s a label that stays the same until you deliberately change it.

Time Dimensions (When?):

  • Date: January 15, 2025
  • Day of week: Monday, Tuesday, Wednesday
  • Month: January, February, March
  • Quarter: Q1, Q2, Q3, Q4
  • Hour: 1 AM, 2 AM, 3 AM

Source Dimensions (Where did they come from?):

  • Campaign: Email_Growth_Promo_Aug_24
  • Ad platform: Facebook, Google, LinkedIn
  • Traffic source: Organic search, Paid search, Direct, Referral
  • Referrer: Google.com, news-site.com, social-platform.com

User Dimensions (Who are they?):

  • Customer segment: SMB, Mid-Market, Enterprise
  • Geography: United States, Canada, UK
  • Industry: Healthcare, Finance, Retail
  • Company size: 1-10 employees, 11-50, 51-200

Product Dimensions (What did they interact with?):

  • Product: Growth product, Pro product, Enterprise product
  • Feature: Payment processing, Reporting, Automation
  • Page: Home page, Pricing page, Blog, Docs
  • Device: Mobile, Desktop, Tablet

Dimensions don’t “improve” or “decline.” They just categorize. A visitor is either from “Organic” or “Paid.” A customer is either in “Enterprise” or “SMB.” These categories don’t change—the distribution of your data across those categories changes.

Why This Distinction Matters

Most founders confuse metrics and dimensions. It sounds pedantic, but it leads to bad analysis.

Example mistake:

  • You look at your data and see “Mobile: 2.5%”
  • You think: “Mobile conversion rate is 2.5%”
  • You panic: “Our mobile conversion rate is terrible. Let’s fix the mobile experience.”

But what does “Mobile: 2.5%” actually mean?

  • If “Mobile” is a dimension and “2.5%” is a metric, then you’re saying: “Of all the mobile visitors we had, 2.5% converted.”
  • If you compare to desktop (3.2% conversion rate), mobile looks worse.

But here’s the question you didn’t ask: Why are the visitors different?

Maybe mobile visitors are coming from paid social ads (which tend to attract lower-intent users). Desktop visitors are coming from organic search (higher-intent). The difference isn’t the device; it’s the quality of traffic.

By confusing the dimension (device) with the metric (conversion rate), you misdiagnosed the problem.

Better analysis:

  • Metric: Conversion rate
  • Dimension 1: Device
  • Dimension 2: Traffic source
  • Question: What’s the conversion rate for mobile visitors from organic search?
  • Answer: 4.1% (actually better than desktop organic at 3.8%)

Now you see: The device isn’t the problem. The traffic source is. Mobile from paid social underperforms because the audience quality is low, not because the mobile experience is bad.

How to Use This in Your Business

When you analyze your data, always ask three questions:

1. What is the metric? (What am I actually measuring?)

  • Conversion rate? Revenue? Customer count? Users per day?

2. What dimensions am I slicing by? (How am I breaking it down?)

  • By traffic source? By geography? By customer segment? By time period?

3. What dimension is actually causing the variance? (What’s really driving the difference?)

  • Is it because of the source, the customer type, the time period, or something else?

Example: Your revenue is down 20% compared to last month.

Initial question: Why is revenue down?

Bad analysis: “Revenue is down. Let’s cut costs.”

Better analysis:

  • Metric: Total revenue
  • Last month: $500k
  • This month: $400k
  • Dimension: Product type
  • Enterprise product: $300k (no change)
  • Mid-market product: $100k (down from $150k)
  • Insight: Enterprise is stable. Mid-market dropped.
  • Next dimension: Customer segment
  • Mid-market customers in “Tech” vertical: $80k (down from $120k)
  • Mid-market customers in “Healthcare”: $20k (no change)
  • Insight: Tech vertical is declining, not healthcare.
  • Next dimension: Time period (day by day)
  • First half of month: $80k (normal)
  • Second half of month: $20k (dropped sharply)
  • Insight: Something happened mid-month in the Tech vertical.

Now you’re asking smart questions: What happened mid-month with Tech customers? Did a competitor launch something? Did we change pricing? Did a key customer churn?

Without breaking down by dimensions, you would have blamed generalized problems (economy, market saturation, product quality). With dimensions, you pinpointed the real issue.

Building Your Metrics vs. Dimensions Framework

We recommend listing out your key metrics and dimensions so everyone on your team uses them consistently.

Key Metrics (for your business):

  • Total revenue (MRR or monthly)
  • Customer count
  • Conversion rate
  • Churn rate
  • Customer acquisition cost
  • Average order value

Key Dimensions (for your business):

  • Time (month, week, day, hour)
  • Traffic source (organic, paid, direct, referral)
  • Customer segment (SMB, mid-market, enterprise)
  • Geography (US, EU, APAC)
  • Product type (if you have multiple products)

By standardizing this, your team stops arguing about what things mean. Everyone knows: “Revenue sliced by traffic source” always means the same thing. “Conversion rate by device” always means the same calculation.

The Takeaway

Metrics are numbers. Dimensions are labels.

You improve metrics. You understand them by dimensions.

If you’re not clear on which is which, your analysis will be confusing and your decisions will be wrong.

We help you establish clean metric and dimension definitions so your data is consistent, comparable, and actually useful for making decisions.