Key Driver Analysis

Key Driver Analysis helps you discover which columns in your data most influence a chosen outcome. It works well for survey and behavioral datasets that mix numbers, opinion scales, and categories.

Open the tool

1. Click the More menu.
2. Choose Find key drivers.

This opens a panel where you pick an outcome and candidate drivers. AddMaple will quickly test likely relationships and then run a Random Forest model to estimate each factor's impact.

How it works (quick version)

• We scan for likely relationships between your selected outcome and other columns.
• We train many small decision trees (a "Random Forest") that handle mixed data types and non‑linear patterns.
• We score each factor by measuring how much model accuracy drops when that factor is temporarily scrambled. Bigger drop ⇒ more important.

Why Random Forest (not just regression)?

Most people have heard of regression. We use Random Forest because it works better for the kinds of mixed, real‑world data you analyze in AddMaple.

• Mixed data with minimal prep: Handles numbers, single/multi‑choice, and Likert scales together without heavy encoding or strict scaling requirements.
• Non‑linear patterns and interactions: Trees naturally capture thresholds and cross‑effects; regression assumes straight lines unless you hand‑craft terms.
• Robust in practice: Less sensitive to outliers and missing values; similar drivers tend to share credit instead of one being over‑fit.
• Clear importance scoring: We report how much accuracy drops when a factor is shuffled. Bigger drop ⇒ more important.

When is regression great? When inputs are mostly numeric and relationships are roughly linear; it gives interpretable coefficients. For mixed survey/behavior data, Random Forest usually gives more reliable driver rankings.

Step 1 — Choose an outcome

Pick one column you want to explain better. Good choices include:

• Numbers (e.g., satisfaction score, time on task)
• Opinion scales (Likert)
• Single categories (e.g., NPS bucket, Yes/No, churned vs active)

Avoid choosing an outcome that will be duplicated among the candidate drivers (data leakage).

Step 2 — Pick possible drivers

AddMaple auto‑suggests columns that look related to your outcome. You can add or remove candidates freely. Useful drivers often include demographics, usage/behavior, attitudes, and experience ratings.

Supported driver types include numbers, opinion scales, single categories, and multi‑select categories.

Tip: Very similar columns can share credit and split their importance. If you see near‑duplicates, consider choosing just one.

Advanced options (optional)

• Number of trees: More trees stabilize rankings but take a bit longer.
• Tree depth: Limits how detailed patterns can get. Lower is simpler and safer.
• Features per split: Fraction of drivers tried at each split to add variety and reduce bias.
• Minimum rows to split: Smallest group size before the tree splits (helps smooth out noise).

You can reset to recommended settings anytime.

Reading the results

After you run the analysis, you'll see:

• Top drivers and a short summary sentence describing what seems to move your outcome
• An importance table: higher scores mean a factor explains more of the differences you see
• Estimated accuracy: the model's out‑of‑bag balanced accuracy on unseen rows

Interpretation tips:

• Associations, not strict causation. Use as directional guidance.
• Very small categories can be noisy — consider combining rare groups or filtering.
• Similar drivers may split importance between them.

Good to know

• Works well with mixed data (numbers, scales, single/multi‑choice)
• Captures non‑linear patterns and interactions automatically
• Robust to outliers and missing values

Availability: Key Driver Analysis is limited to certain plans.