KPIs, Pivots & Measures

📊 KPIs, Measures and Pivots

GermainUX provides preconfigured KPIs, measures and pivots for analyzing user experience, business processes, marketing, web activity and technology performance. Organizations can also create and customize them for their own applications and business requirements.

Together, these three elements answer three different questions:

For example:

Analyze the 95th-percentile duration (measure) of User Clicks (KPI), broken down by application page (pivot).

This model allows the same monitored data to support high-level dashboards, segmented comparisons, detailed drilldowns and root-cause analysis.

🎯 Business Outcomes

KPIs, measures and pivots help organizations:

• Monitor adoption, conversion, productivity, user experience and technology health.

• Define consistent calculations across dashboards and teams.

• Break down performance by users, applications, workflows, products and other dimensions.

• Identify outliers, degraded segments and leading factors.

• Move from a high-level result to the instances and evidence behind it.

• Create custom analytics without building separate reporting logic for every use case.

• Apply SLAs, alerts and automation to the conditions that matter.

🔗 How the Three Elements Work Together

A GermainUX visualization begins with a KPI and then applies a measure and, when needed, one or more pivots.

KPI: What happened?
        ↓
Measure: What value should be calculated?
        ↓
Pivot: How should the value be segmented?
        ↓
Dashboard, analysis, alert or report

Examples include:

📈 KPIs

A KPI defines the type of fact, event, transaction or business condition to be monitored and analyzed.

Depending on the use case, a KPI can represent:

• A user interaction

• A user session

• A business-process instance

• A journey milestone or drop-off

• A conversion or abandonment

• An error or exception

• An application transaction

• A performance or availability event

• A marketing or campaign outcome

• A custom business event

KPI configuration can include:

• Fact or event type

• Default measure

• Available pivots

• SLAs and statistical SLAs

• Filters and constraints

• Relationships with other KPIs

• Drilldown and analysis behavior

• Alerts and automation

🗂️ Preconfigured KPI Categories

🔢 Measures

A measure defines the numerical value or calculation applied to the instances of a KPI.

Measures can include raw numeric attributes and calculated aggregations such as:

• Count

• Sum

• Average

• Minimum and maximum

• Median

• Percentiles

• Rate or percentage

• Duration

• Throughput

• Error frequency

• Conversion rate

• Lost productivity

• Revenue or business impact

💡 Why the Measure Matters

The same KPI can produce very different insights depending on the selected measure.

For a User Click KPI:

Measures should be selected according to the decision being made. Averages are useful for overall trends, while percentiles reveal experiences that an average can conceal. Counts show scale, while rates make segments of different sizes comparable.

🗂️ Preconfigured Measure Categories

🔄 Pivots

A pivot is a dimension used to group or segment KPI results.

Pivots help answer where, who, which or under what conditions a result occurs.

Examples include:

• Application and environment

• Page, screen or feature

• User, role or team

• Business process and milestone

• Customer segment

• Product, service or campaign

• Device, browser or operating system

• Application version

• Country, region or location

• Error category or technology component

📊 Pivot Dashboard

The Pivot Dashboard automatically presents the selected KPI and measure across multiple relevant dimensions. This helps teams identify where a problem is concentrated without configuring every breakdown manually.

GermainUX preconfigures at least ten pivots for applicable KPIs. The exact pivots available depend on the KPI, data model, monitored application, product version and collected attributes.

Pivots can reveal:

• The pages with the slowest interactions

• The teams experiencing the most lost productivity

• The products contributing the most abandonment

• The application versions generating the most errors

• The devices or regions with the lowest conversion

• The workflow stages where completion degrades

🗂️ Preconfigured Pivot Categories

Pivots Versus Filters

Pivots and filters serve different purposes.

A filter reduces the analysis scope. A pivot keeps the relevant groups visible so they can be compared.

Using Multiple Pivots

Multiple pivots can provide a more precise breakdown.

For example:

User Click duration by application page, then by browser version.

This can reveal that a page is slow only for one browser version. However, excessive pivot combinations can create many small groups and misleading results. Always consider the number of instances behind each segment.

From Aggregate to Root Cause

KPIs, measures and pivots support the complete GermainUX analytics workflow:

1. Aggregate Dashboard: Monitor the KPI and selected measure over time.

2. Pivot Dashboard: Identify the dimensions where the result is concentrated.

3. Drill-through Dashboard: Review the relevant users, sessions or transactions.

4. Analysis Dashboard: Compare baselines, related KPIs, Leading Factors and individual evidence.

5. Session Replay or Trace: Understand what happened in a specific instance.

This allows teams to move from a high-level symptom to evidence-backed root cause without losing the original analytical context.

Create Custom KPIs, Measures and Pivots

Organizations can extend the preconfigured analytics for application-specific and business-specific needs.

Custom configuration can define:

• New KPI and fact types

• Custom numerical attributes and calculations

• Organization-specific pivot dimensions

• Names, descriptions and display formats

• Default measures and pivots

• Filters, constraints and relationships

• SLAs and statistical thresholds

• Drilldown and analysis behavior

• Alerts, reports and automated actions

Custom elements should use consistent definitions and units so results remain comparable across dashboards and teams.

Analytical Best Practices

• Choose the KPI that represents the event or outcome being investigated.

• Use a measure aligned with the business question.

• Compare counts and rates when segment sizes differ.

• Use percentiles when averages may hide slow or poor experiences.

• Verify the number of instances behind each pivot value.

• Avoid conclusions from very small segments.

• Apply filters to remove irrelevant data before comparing pivots.

• Drill into representative instances before deciding on root cause.

• Validate that custom formulas and units are consistent.

Get More Information

Contact GermainUX Support for help selecting or creating KPIs, measures and pivots for your use case.