15 April 2026
CX analytics teams operating across insurance firms don’t just have to keep customers happy. They also need to maintain compliance with stringent regulations.
Our webinar on the benefits and drawbacks of GenAI in CX analytics explored what this looks like in practice. Relative Insight CRO James Cuthbertson shared how a team at an insurance company could pivot to using GenAI in its CX program — and why this wouldn’t enhance the process.
James outlined that the manager of this small, but capable, analytics team is responsible for understanding why CSAT scores rise and fall, where journeys break down and what’s driving churn.
The wrong AI for the job
To do this, the team has adopted GenAI platforms like Copilot. This has enabled the analysts to create insight reports faster than ever before, hitting their briefs week on week.
Using GenAI has enabled them to produce summaries of call center transcripts, claims feedback, complaints data, digital journeys and more. On the surface, this looks like progress: faster outputs, broader coverage, more ‘answers’.
However, greater output isn’t leading to better outcomes. As the CX team manager reviews the reports, there are clear inconsistencies. One week, a key driver of dissatisfaction is claims processing delays, yet the next week it’s communication gaps.
When the managers asks team members how they reached these conclusions, and whether they can repeat the process, they’re unable to do so. The insights are generated through LLMs, and are therefore non-deterministic.
The same query can produce different outputs, making the ‘working’ behind the insight is nonexistent.
Diminishing returns of GenAI CX reporting
This lack of clarity and defensibility creates problems in the organization, rather than solving them. Due to these overlapping – sometimes conflicting – signals, the CX manager hesitates to take insights to senior leadership because they lack confidence in their reproducibility.
In turn, the analysts grow frustrated. They’re hitting the brief and using the best tools available to them. However, their work isn’t landing.
When findings are eventually filtered up to decision makers, they invariably ignore the outputs. The insights don’t have the metrics, evidence or auditability to convince them to make changes that require budget and need to tested for compliance.
This means that despite having more data and more analysis than ever, the business moves slower in acting on that data.
The shift to deterministic AI
This is where Relative Insight reframes the problem. Instead of generating interpretations, it applies deterministic analytics to customer feedback across every touchpoint. This pinpoints exactly what’s driving changes in KPIs.
Using Relative Insight’s deterministic engine creates auditable, metric-backed insights. They are consistent, reproducible and fully traceable — meaning that CX leaders can confidently present insights to decision makers that are grounded in evidence.
Relative Insight’s platform shifts internal conversations. Instead of asking analysts to defend how they arrived at an answer, the answer itself is inherently defensible. Every insight can be traced back to specific words, phrases and themes in customer feedback.
Weekly reporting becomes structured and repeatable. This consistent format is something that leadership can trust and act on.
An analytics team that drives action
Incorporating deterministic AI gives the small analytics team some internal momentum. Insights are not longer ignored. They’re rooted in evidence and they’re defensible, meaning that decision makers use them.
Leadership teams take them into board discussions. Operational teams act on them to refine claims processes, improve communications and redesign key journey touchpoints. Compliance teams are confident that changes are being made for the right reasons.
The environment has completely changed for the team manager. Now they have a high-performing team whose insights are driving organizational change. Rather than spending time asking team members to defend their insights, the manager can get on with higher value work.
In a sector like insurance where decisions carry financial, regulatory and reputational weight, using deterministic AI, not just GenAI, is a non-negotiable.
Want to find out how Relative Insight can help your team scale your analytics outputs while maintaining rigor? Book a call with our team.
