AI and personalisation at scale: reshaping customer engagement in insurance
Customer expectations keep rising. Price and coverage matter, but experience decides loyalty. The shift underway is clear: move from transactional interactions to connected, individualised engagement across every touchpoint.
Insurers across the UK, US, and Asia are pushing beyond basic segmentation. They're building real-time, AI-enabled journeys that treat each customer as an individual, whether the interaction is digital or with an agent. Leaders like Aviva, Allstate, and Ping An show what this looks like in practice.
IAG's digital pivot: from systems of record to systems of engagement
Australia's IAG offers a practical example. Its partnership with Adobe and investment in a retail enterprise platform signals a shift to an integrated system of engagement - a layer that connects data, channels, and decisions across the journey.
"It's a platform that allows us to improve and uplift the way we engage with our customers through our digital channels in particular, but also through our marketing channels," said Nandor Locher, IAG's executive general manager of digital business. "It's more a system of engagement rather than a system of record and builds on top of the enterprise platform that we have built."
This choice matters. Instead of stitching together point solutions, IAG is using a unified platform to connect signals, context, and actions. "It allows us to engage with our customers in a much more personalised way, really understanding their interaction behaviours across the different channels and start connecting those experiences across the different touchpoints right along the customer journey," Locher said.
For context on this approach, see Adobe Experience Platform, which centralises profiles, decisioning, and activation across channels. Learn more.
Personalisation at scale means connected, real-time, and human-aware
The target isn't one-off campaigns. It's consistent, relevant experiences across millions of interactions - all informed by live data. "This is a big step forward to really do this personalisation at scale across all touchpoints," said Locher. "It's highly automated and really down to an individual rather than just at the segment level."
"The big word is the connected experience where the different touchpoints in previous stages are providing context for the current conversation that a customer may be having," he said. That includes the call centre. Agents see journey context, previous actions, and predicted next best steps - not just policy details.
AI is central here. "AI is increasingly built into this platform… allowing us to focus on the insurance business and the customer engagement and leverage these capabilities," Locher said. Better real-time information is improving both growth and efficiency.
Integrated platform vs best-of-breed: what insurers should weigh
- Speed to scale: Integrated platforms reduce integration debt and time to value. Best-of-breed can win on depth but slows orchestration.
- Data unity: A single identity and profile store is essential for consistent decisioning. Fragmented stacks risk conflicting treatments.
- Change management: One platform simplifies training and governance. Multiple tools increase operational complexity.
- Total cost: Consider build, run, and upgrade costs - not just licenses. Long-term maintenance often outweighs initial savings.
What "good" looks like: the core capability stack
- Unified profile: Consent-aware identity resolution with streaming updates.
- Decisioning brain: Next-best-action and eligibility using rules plus machine learning.
- Journey orchestration: Cross-channel triggers, throttling, and frequency controls.
- Content and offers: Modular content, dynamic pricing/coverage offers, and clear compliance guardrails.
- Agent assist: Real-time context and guidance in the contact centre and broker portals.
- Measurement: Holdout groups, incremental lift, and closed-loop learning into models.
- Risk and governance: Model approvals, bias checks, audit trails, and strict data minimisation.
Where to start: a 90-day playbook
- Choose one journey with volume and value: quote-to-bind, renewal, or first notice of loss.
- Unify the signals: web/app events, policy and claims data, consent, and call-centre outcomes.
- Define policies: eligibility, fairness constraints, channel prioritisation, and contact frequency caps.
- Ship an MVP: two to three next-best-actions, one digital treatment, one agent assist use case.
- Measure what matters: conversion, retention, cost-to-serve, and time-to-resolution. Use holdouts.
- Iterate weekly: feed outcomes back into models and rules; expand to the next touchpoint.
Compliance, trust, and control
- Consent-first design: honour preferences across every channel and partner.
- Transparent decisioning: record "why" for offers and treatments; make explanations available to agents.
- Bias and fairness: monitor model drift and segment outcomes; set thresholds for interventions.
- Security and privacy: minimise data, expire it quickly, and maintain full auditability.
Operating model: teams that ship
- Cross-functional pods: product owner, marketer, underwriter, data scientist, engineer, and compliance.
- Shared KPIs: tie incentives to customer outcomes and efficiency, not channel silos.
- Agent feedback loop: capture frontline insights and bake them into decisioning and playbooks.
The takeaway
The market is moving to connected, individualised engagement - at scale and in real time. IAG's path shows the benefit of an integrated platform that fuses data, decisioning, and delivery. Insurers that master this will win on loyalty, growth, and cost-to-serve.
If your teams need structured upskilling to execute on this, explore role-based programs at Complete AI Training.
Your membership also unlocks:
