Go beyond correlation to answer the question that matters most: "If I improve this driver, what return will I get?"
A driver can be highly correlated with satisfaction but offer low returns if you're already performing well. Traditional analysis can't distinguish between "critical" and "already excellent."
A driver might show modest correlation but offer exceptional ROI because current performance is low. Traditional analysis would deprioritize this high-value opportunity.
Not all important drivers are created equal. Positive drivers create gains when improved; hygiene factors destroy satisfaction if they decline but don't help if improved.
Catalyst combines importance, impact, and performance to create a complete investment picture
"How well does this driver predict the outcome?"
Machine learning captures non-linear relationships that traditional correlation misses. Output: Importance Score (0-100).
"How much outcome change results from improving this driver?"
Simulates improving each driver while holding others constant. Reveals which drivers have steep slopes vs. diminishing returns. Output: Impact Index (0-100).
"Where do you currently stand on each driver?"
Low performance + High impact = Immediate opportunity. High performance + High importance = Protect but don't over-invest. Output: Performance Score (0-10).
Clear, actionable categories that tell you exactly what to do
Your top investment priorities. These drivers strongly predict the outcome and deliver exceptional ROI when improved.
Underperforming drivers with immediate ROI opportunity. Low-hanging fruit that competitors likely also miss.
Critical foundations that destroy satisfaction if they decline but don't create upside if improved.
You're already performing well, so further investment yields diminishing returns.
Limited value creation potential regardless of investment level.
Most CX programs rely on stated importance to prioritize action. But what if the attributes customers rate as important aren't actually the ones that drive their behaviour? We built a synthetic retail banking dataset to demonstrate how the Catalyst Framework reveals hidden priorities.
→ Invest heavily. Customers understate its importance, but it's the #1 driver of NPS. Every 1-point improvement generates 2.1× more value than Mobile App.
→ High-impact opportunity. Low current performance + high behavioural impact = immediate ROI. Service recovery is a loyalty multiplier.
→ Maintain, don't over-invest. Traditional analysis flagged this as #1 priority. But you're already excellent — further investment yields diminishing returns.
→ Protect, don't improve. High importance but negative direction — decline destroys satisfaction, but improvement doesn't create upside. Maintain current levels.
To demonstrate the Catalyst Framework, we used a synthetic dataset designed to reflect realistic South African retail banking customer experience dynamics. This is not actual FNB customer data — it's a simulation built to illustrate how the framework reveals hidden priorities.
The synthetic dataset was constructed using:
This approach mirrors how the Catalyst Framework would be deployed on actual client data. Knowsis offers synthetic data generation services for analytical prototyping, concept validation, and scenario planning — allowing you to test frameworks before committing to full research investment.
Stop guessing where to invest. Get clear, defensible, ROI-focused priorities that drive measurable business outcomes.
Start Driver Analysis