KPI Relationships

ServiceAnalytics

Availability2014.1 or later

Relationship Types

Kinds of linked KPIs

Configuration

Set up linked KPIs

Pivot a Related KPI

Segment a trend relationship

Configuration Recommendations

Best practices for linked KPIs

Overview

KPI Relationships define how KPIs are related to each other and help GermainUX identify signals that may explain changes in a KPI being analyzed. Relationships are used by the Analysis Dashboard to prioritize related KPIs and provide additional context during analysis and root-cause investigation. For example:

Source KPI	Related KPI	Relationship
User Click Failure	Mouse Click Error	Mouse Click errors may contribute to failed user interactions.
Order Sync Issue	Salesforce API Error	API errors may provide context for synchronization issues.
Shopify Conversion Rate	Checkout Page Load Time	Increasing page load time may be associated with decreasing conversion.

Relationships are directional. A relationship configured from KPI A to KPI B does not automatically create a relationship from KPI B to KPI A.

How KPI Relationships Support Analysis

When a KPI is opened in the Analysis Dashboard, GermainUX evaluates its configured relationships to identify signals that may help explain the KPI’s behavior. Related KPIs can be prioritized using factors such as:

Deviation from baseline
Direction of movement
Relationship strength
Relationship type
Co-movement with the KPI being analyzed

This helps investigators quickly identify KPIs that:

May contribute to the observed issue
Move with the analyzed KPI
Move in the opposite direction
Provide useful contextual events
Execute within the analyzed KPI’s activity

Instead of manually comparing multiple dashboards and KPIs, investigators can focus on the signals most relevant to the current analysis.

Relationship Types

Type	Meaning	Example
Contributor	Changes or issues in the related KPI may contribute to changes in the source KPI.	Mouse Click errors contributing to User Click failures
Correlated	The KPIs tend to move together without asserting that one causes the other.	API latency and page response time increasing together
Contextual	The related KPI provides useful investigative context.	A deployment event occurring around the time of an error spike

A relationship helps prioritize an investigation; it does not, by itself, prove causation.

Configure a KPI Relationship

The easiest way to create a relationship is through the Add KPI Relationship wizard from the source KPI’s configuration. Germain Workspace > Left Menu > Wizards > Add KPI Relationship

Open the KPI from which the relationship should originate. This is the source KPI.
Open KPI Relationships.
Launch the Add KPI Relationship wizard.
KPI Relationship Wizard - GermainUX
Select the related KPI.
Configure the relationship.
Select Finish or Save.
Confirm that the relationship appears under the source KPI’s KPI Relationships.

Configuration

Field	Required	Description
Name	Yes	Unique name identifying the relationship and related KPI.
KPI	Yes	KPI related to the current source KPI.
Type	Yes	Contributor, Correlated, or Contextual.
Direction	Yes	Expected direction of movement: Same or Inverse.
Strength	Yes	Expected relationship strength: Strong, Moderate, or Weak.
Constraint	No	Filter that must be satisfied for the relationship to apply, such as `userError == true`.
Field	No	Field used to correlate instances of the two KPIs, such as `sessionId`.
Measure	No	Measure evaluated for the related KPI, such as `avg(duration)`. If omitted, GermainUX uses the KPI’s default numerical property or count when no property is available.
Show As	Yes	Displays the related KPI as a Trend or Marker in the Analysis Dashboard. Trend is the default.
Child	No	Indicates that the related KPI executes within the source KPI’s context and supports nested activity in the Analysis Trace Timeline.

Direction

Direction	Meaning	Example
Same	The related KPI is expected to move in the same direction as the source KPI.	Error Clicks and failed User Clicks increase together.
Inverse	The related KPI is expected to move in the opposite direction.	Page load time increases while conversion decreases.

Direction describes the expected numerical relationship, not whether the behavior is good or bad.

Strength

Strength	Use When
Strong	The relationship is consistently expected and highly relevant to investigations.
Moderate	The relationship is meaningful but may not appear in every situation.
Weak	The relationship is indirect, occasional, or exploratory.

Strength affects prioritization in the Analysis Dashboard. Configure it based on observed behavior and domain knowledge rather than the importance of either KPI.

Example

Assume Mouse Click errors are a leading contributor to User Click failures occurring within the same user session. Configure the relationship on the User Click KPI:

Field	Value
Name	`MouseClickContributor`
KPI	`Mouse Click`
Type	`Contributor`
Direction	`Same`
Strength	`Strong`
Field	`sessionId`
Show As	`Trend`
Child	Enabled

To include only Mouse Clicks associated with an error, add a constraint such as: userError == true Configured KPI Relationships become part of the investigation in the Analysis Dashboard.

Depending on the relationship, a related KPI can be displayed as a Trend or Marker.

Trend

Use Trend when changes over time, deviation from baseline, or co-movement are important to the investigation. A related KPI can display:

Current value
Comparison with baseline
Trend
Deviation or health
Co-movement with the analyzed KPI

Select a related KPI’s trend to add it to the main chart and compare it directly with the KPI being analyzed.

Marker

Use Marker for discrete events where timing is more important than an aggregated trend. Examples include:

Errors
Crashes
Deployments
Configuration changes

Markers are displayed on the Analysis Dashboard timeline. Select a marker to investigate the individual occurrence and compare its timing with changes in the analyzed KPI.

A Trend relationship can be segmented using a Pivot to determine whether the relationship is concentrated within a particular population, application, technology, or environment.

Select Pivot By on the related KPI.
Pivot By - GermainUX
Select a dimension such as Country, Browser, or Server.
Review the resulting segments.
Pivot values - GermainUX
Compare their health, measure, count, and trends.
Select a segment to add it to the main chart.

This can reveal, for example, that a performance issue associated with declining conversion affects only a particular browser, application version, server, geography, or user population.

Filter Markers

When Marker relationships are displayed, use Filter Events to control which events appear on the Analysis Dashboard timeline.

You can:

Show or hide individual Marker KPIs
Filter markers by field values

Search for KPIs or values
Select or clear all markers

Changes are applied immediately to the visible timeline.

KPI Relationships and Root-Cause Analysis

KPI Relationships provide one source of evidence for Root-Cause Analysis. A typical investigation can progress from:

KPI ChangeRelated KPIsAffected PopulationIndividual InstancesTrace / Session ReplayLikely Root Cause

Relationships help identify where to investigate first. The underlying events, transactions, traces, Session Replay, and other monitored evidence help determine what actually occurred. For example: Conversion decreases
→ Checkout Page Load Time appears as a strongly related KPI
→ Pivot identifies the affected browser population
→ Individual sessions are investigated
→ Session Replay and related transactions provide supporting evidence
→ Likely cause is identified This allows KPI Relationships to connect high-level business or experience changes with the detailed evidence available in GermainUX.

Configuration Recommendations

Configure relationships on the KPI users will investigate as the source.
Use Contributor only when there is a credible contributing mechanism.
Use Correlated when KPIs move together but causation has not been established.
Use Contextual for events or signals that help explain the timeline.
Choose Same or Inverse based on expected numerical movement.
Use a correlation field such as sessionId when instance-level association matters.

Show all 11

Apply constraints to exclude irrelevant instances.
Select a Measure that matches the analytical question.
Use Trend for continuous or aggregated signals.
Use Marker for discrete events.
Validate relationships against historical examples before assigning Strong strength.

Root-Cause Analysis

Analytics

More Analytics Features

KPIs, Pivots & Measures

Manage KPIs

Contact GermainUX Support

The GermainUX team helps you get up and running faster, empower your team, turn data into actionable insights, and optimize performance to deliver measurable business outcomes.

Can't find what you need?

Search all GermainUX docs with Ctrl + K, or use the feedback buttons at the bottom of this page to tell us what is missing.

The content of this document is proprietary to GERMAIN SOFTWARE.

GermainUX Documentation

Relationship Types

Configuration

Pivot a Related KPI

Configuration Recommendations

Overview

How KPI Relationships Support Analysis

Relationship Types

Configure a KPI Relationship