In This Article

Overview

Start Here: Identify the Symptom

Key Point to Remember

What Happens Next

Next Steps

Overview

This article helps you diagnose and resolve common issues encountered when running analysis in Value Hound. It focuses on problems related to analysis setup, data availability, and interpretation of outputs.


Structural and KPI configuration issues are covered in Troubleshooting Value Driver Trees.

Start Here: Identify the Symptom

Before troubleshooting, identify what you are seeing:

  • No chart or table appears after running analysis
  • Analysis runs but shows empty or zero results
  • Results look extreme or counterintuitive
  • Attribution shows 100% or overlapping contributions
  • Variance impacts do not add up as expected

Once you identify the symptom, use the relevant section below.

1. No Analysis Results Are Displayed

Common symptoms

  • Chart area is blank
  • No values appear after selecting Run
  • Results do not update when inputs change

What to check

  1. KPI values exist: Confirm that all KPI nodes in the tree have values for the selected field and period.
  2. Field selection: Ensure the selected base, control, or analysis fields (Actual, Target, Baseline) contain data.
  3. Tree calculation resolves: Confirm that the Value Driver Tree outcome node displays a value before running analysis.
  4. Analysis has been run: Ensure Run (or Auto Run) has been triggered after changing settings.

Typical fixes

  • Switch to a view with populated fields
  • Add missing KPI values
  • Re-run the analysis

2. Analysis Results Look Empty or Zero

Common symptoms

  • All bars are zero
  • Outcome difference is zero
  • Sensitivity shows no variation

What to check

  1. Control and analysis fields are different: In Attribution Analysis, if Control and Analysis fields are the same, the result will be zero.
  2. Variance inputs applied: In Variance Analysis, confirm that non-zero variances have been entered.
  3. Uniform KPI values: If all KPIs have identical values across fields, there may be no difference to analyse.

Typical fixes

  • Select different control and analysis fields
  • Enter meaningful variances
  • Verify KPI values differ between fields

3. Results Look Extreme or Unexpected

Common symptoms

  • One KPI dominates Sensitivity or Attribution
  • Very large positive or negative impacts
  • Small changes produce large swings

What to check

  1. Multiplicative logic: Identify calculations that multiply values (for example, rate × time × price).
  2. Upstream drivers: KPIs high in the tree can dominate analysis results.
  3. Zero or near-zero values: Targets or baselines set to zero can cause extreme impacts.

Typical fixes

  • Validate KPI target and baseline values
  • Review calculation structure
  • Use Variance Analysis to sanity-check individual impacts

4. Attribution Shows 100% or Overlapping Contributions

Common symptoms

  • One KPI shows 100% contribution
  • Multiple KPIs show contributions exceeding 100% in total

Why this happens

  • Attribution assigns overlapping explanatory impact to KPIs at the output node. In non-linear or multiplicative trees, multiple KPIs can each explain a large portion of the same outcome difference.
  • This is expected behaviour, not an error.

What to check

  • Confirm the outcome difference being analysed
  • Review which KPIs changed between fields
  • Validate model structure and leverage points

5. Variance Impacts Do Not Add Up

Common symptoms

  • Combined variance results are smaller than expected
  • Individual variance impacts do not sum linearly

Why this happens

  • Variance impacts are not additive when KPIs interact through shared calculations or constraints.

What to check

  • Whether multiple KPIs affect the same part of the tree
  • Whether upstream KPIs limit downstream effects

6. Results Differ Between Users

Common symptoms

  • Two users see different analysis results
  • Charts differ despite using the same tree

What to check

  1. Access and position context: Users may be analysing different data scopes based on their position.
  2. View and field selection: Confirm both users are using the same view and analysis settings.
  3. Recent data changes: Check KPI audit logs for recent updates.

7. When Analysis Results Change Unexpectedly

If analysis outputs change without rerunning the analysis:

What to check

  • KPI value updates
  • Idea changes affecting inputs
  • Tree edits made by other users

Remember:

  • Analysis reflects the current state of data and structure
  • Exported analysis results do not update automatically

Key Point to Remember

Value Hound analysis reflects the current data and model state. Unexpected results are usually caused by data changes, field selection, or model structure — not calculation errors.

What Happens Next

If issues persist:

  • Re-run analysis with simplified settings
  • Validate KPI values and tree structure
  • Use Variance Analysis to isolate effects

For interpretation guidance, refer back to:

Next Steps

To learn about frequently asked questions, see: