Growth Analytics Design · Metric Set
Designing "Growth Analytics" for Small Teams: A Metric Set to Track by Business Area
Small teams often face the problem of “too many metrics to track.” That’s why it’s important to define metrics not as something to “collect,” but as a “design” that directly supports decision-making. Below is a metrics set organized by business area—what to measure, at what level of detail, and when to review it.
1) Assumptions for Indicator Design: North Star KPI and Observation Layer
First, we select one North Star KPI. Next, we break that KPI down into three layers of observation.
- Value Layer：What happened to the customer (e.g., retention rate, realization of value from use)
- Earned layer：Why it happened (e.g., number of acquisitions, CVR, onboarding reach)
- Productivity layer: What moved the team forward (e.g., lead time per initiative, number of validation cycles)
In a small team, it’s more realistic to “build learning over a weekly cadence” than to “watch everything every day.” We also design the cadence: the value layer is weekly, acquisition is daily, and initiatives are weekly.
2) By business area: The KPI set to watch
Product/Subscription (SaaS, B2B)
- Value：Retention rate (logo/revenue), utilization value action reach rate
- Get: Trial start → activation CVR from reaching effective use
- Productivity：time to introduce initiatives, improvement cycle (hypothesis → testing → implementation)
Tip: Manage cancellation reasons by “patterns” rather than “numbers,” and map them to improvement themes.
Outsourced / development (project revenue)
- Value：Case initial scope retention rate, post-delivery additional order rate
- Get: lead-to-meeting-to-proposal-to-order conversion rate
- Productivity: Estimate lead time, rate of re-quoting occurrences, quality metrics (rework)
Note: Stage conversion rate is measured by the “bottleneck,” not by the “responsible department.”
EC/Community-based (continuous purchasing-relationship)
- Value: repeat purchase rate, purchase frequency, and reactivation rate after dormancy
- Get：First-time purchase CVR, drop-off rate before first-time purchase
- Productivity: The speed at which campaign improvements are reflected, and results per delivery
Point: Don’t judge only by short-term results from coupons and ads—connect to indicators of repeatable performance.
Data/Consulting type (performance-based, onboarding support)
- Value: Achievement rates for effectiveness indicators, contract renewal rates
- Earned：PoC→Contract conversion rate, Completed implementation rate
- Productivity: proposal quality (win rate), preparation time, and average implementation effort
Point: The key to effective performance metrics is knowing “when to measure.” Fix the measurement timing.
3) Operating rules for changing KPI to "Action"
Indicators alone won’t move you forward. We’ll standardize the following three:
- Threshold：Set the criteria for upward and downward deviations to clarify when notifications are triggered
- Accountability and Verification：Assign a “verifier” for the hypothesis, not a “handler” for the numbers
- Learning log：For each improvement theme, record what you tried and what you learned
This design shifts decision-making from instinct to data. What’s needed isn’t more reports, but reproducible decision frameworks.
Next step: Create a KPI map by area
In every area, breaking down “Polaris KPI → value/attainment/productivity” and reviewing it on a weekly basis makes the feeling of growth change.
Additional note: signs that analysis is getting “heavy”
If this situation continues, the indicator set may be excessive.
- The time to build the dashboard is exceeding the time for testing the hypothesis.
- The cause of the issue cannot be identified, and in the end it becomes nothing more than “changing the approach.”
- The review ends with a “report,” and the next steps aren’t decided.
To make the design lighter, prioritize metrics that connect to the value layer, and progressively narrow down everything else.