Unlocking Key Metrics: How Operations Designers Use Funnels, Dimensions, and Measurements

As operation designers we often deal with data terms such as PV, UV, conversion rate, click‑through rate, exposure, retention, DOSU, DAU, WAU, etc. Data metrics are the lifeblood of operation design; in an outcome‑driven era, how should we use these data?

So Many Data, How to Identify Key Metrics?

When the amount of collected data grows, it is easy to get "lost in data". Facing a flood of visual numbers like PV and UV, you may struggle to find the key indicators that drive business growth. Although each company and department has different metrics, two methods can help extract key metrics.

Define User Behavior Paths

Set a conversion path such as (Home → Section A → Participate → Download). The number of users who complete the conversion within a given period is counted. This goes beyond traffic analysis and delves into user behavior.

Establish Key Metric Combinations

Good metrics are simple, easy to understand, and often expressed as a ratio or a combination of ratios.

Beyond Data, Understand the Funnel Principle

From the funnel (pyramid) diagram, analyze the loss rate at each layer to quickly locate problems. For example, low activity may indicate a channel issue. By tracing the data upward, you can swiftly address the root cause. The process flows bottom‑up, while verification proceeds top‑down.

Data Validation Analysis: Where to Start

Measurement: Quantitative values such as page views or session duration. Often called "metrics", but some contexts distinguish metrics as calculated results (e.g., total page views divided by total visitors to obtain average views per visitor). In practice, the terms are interchangeable.

Dimension: The perspective from which we view data. For instance, page views (PV) can be examined by date, traffic source, or user cohort. Frequently, multiple dimensions are combined for deeper insight.

How to Distinguish Them?

Dimensions have enumerable members that remain relatively stable. For example, the date dimension includes a finite set of days (365 per year), and the city dimension lists distinct locations.

Explain in More Detail

Measurement: Sometimes called derived metrics when two metrics are combined (e.g., Metric A ÷ Metric B = Metric C). Also note additive vs. non‑additive measurements: UV (unique visitors) is non‑additive because the same user may appear on multiple days, so daily UVs cannot be summed directly.

Dimension: Some dimensions are independent (city, time), while others have hierarchical relationships (province → city, industry → category, grade → class). Hierarchical dimensions enable drill‑down analysis, moving from coarse to fine granularity.

Case Study: Real Environment

From the navigation layout, the expected flow was a step‑wise funnel (Task Page A → Eat Turkey Page B → Prize Page C). In reality, the flow behaved like a waterfall, with a very low proportion of users completing all pages; most dropped off at Page A. This issue is serious and needs resolution.

Applying the funnel principle, the secondary page back‑flow problem requires verification from the top down to locate the source, questioning whether the navigation or primary page content is problematic.

We tried the following improvements:

Highlight the back‑flow button to strengthen guidance.

Hide content to create mystery.

Add interactive effects to make the whole activity more engaging.

Resulting PV ratios for the three stages and the prize page were 1.8 : 1 : 1.7 : 1.3. Compared with the previous tab‑switch format, traffic distribution across sub‑pages improved by 4‑10 percentage points.

Case Study: Choose a Husband?

When Xiao Bei met Jack Ma online, a dilemma arose: who should be chosen as a husband? This whimsical example also illustrates dimensions and measurements.

Dimension: Person Members: Jack Ma, Xiao Bei Measurement: Wealth, appearance, and each factor’s weighted index.

Ultimately, the choice becomes a playful paradox: Xiao Bei as husband, Jack Ma as father.

Conclusion

For operation designers, strong data collection and analysis skills greatly aid growth. By understanding the problems encountered in operations, using the funnel principle to pinpoint issues, and employing dimensions and measurements to validate data, designers can enhance visual appeal, optimize interaction and user experience, and boost activity effectiveness and key metrics.