Comarch User Group 2026 shows loyalty shifting to AI and psychology

At a glance

What this is: Comarch’s event recap says loyalty strategy is shifting from generic discounting toward behavioral psychology, AI-enabled empathy, and fine-grained segmentation tied to measurable financial outcomes.

Why it matters: That matters because identity-adjacent customer decisioning increasingly depends on governed data quality, traceable audience logic, and controlled personalization across programmes that influence access to offers and benefits.

By the numbers:

• Comarch User Group 2026 included 22 loyalty-focused sessions.
• The Loyalty Stream brought together 107 brand strategists from 33 countries.

Context

Comarch’s event recap is about where loyalty programme design is heading, not about a single product feature. The article argues that generic discounts are losing relevance as brands lean harder into behavioral psychology, AI-supported empathy, and more precise segmentation.

For identity and access practitioners, the relevance is indirect but real. Customer decisioning systems increasingly use richer personal and behavioural signals, which means data quality, governance, and auditability matter more as these programmes start making differentiated offers at scale.

Key questions

Q: How should teams govern AI-driven loyalty segmentation when customer treatment changes by behaviour?

A: Treat AI-driven loyalty segmentation as a governed decision layer, not a creative campaign tool. Set approval rules for segment creation, limit which behavioural inputs can influence offers, and keep a clear audit trail for overrides. The goal is to make customer treatment explainable, reversible, and measurable before the logic is scaled across channels.

Q: Why do behavioural loyalty programmes need stronger data governance than points-based schemes?

A: Behavioural programmes rely on finer customer signals, more frequent updates, and more decision paths than points-only models. That increases the chance of bad data, inconsistent segmentation, and hidden bias in offer selection. Stronger governance is needed because the programme is making more consequential decisions with more inputs.

Q: What are the warning signs that micro-segmentation is creating governance drift?

A: Common warning signs include too many exception rules, regional teams using different customer definitions, and no clear link between segment changes and financial outcomes. When the same customer can be treated differently across systems without a documented reason, the segmentation model has drifted beyond useful control.

Q: When should organisations centralise loyalty decisioning instead of letting local teams adapt offers?

A: Centralise when local adaptation starts to weaken consistency, auditability, or financial measurement. Local teams can still add value with contextual knowledge, but they need clear boundaries. If regional discretion changes customer eligibility or pricing without oversight, the governance model is too loose for scale.

Technical breakdown

Behavioral segmentation in loyalty decisioning

Behavioral segmentation uses observed habits, patterns, and context to decide which offer, message, or reward a customer sees. In modern loyalty programmes, that logic can sit across multiple systems, from mobile apps to customer data platforms, making the decision path harder to trace than a simple points model. The security and governance issue is not just personalisation. It is whether the programme can explain why one customer was placed in one segment and another was not, especially when local teams can override central logic.

Practical implication: require traceable segment logic and documented decision inputs for any loyalty workflow that changes customer treatment.

AI empathy and customer-state inference

AI empathy in the article means using behaviour changes as signals that the customer’s situation has changed, so the programme adjusts tone, recommendations, or offers. That creates a governance challenge because the model is inferring a sensitive state from weak indicators. The risk is misclassification, overreach, or tone-deaf automation if training data is poor or business rules are not constrained. This is not just a marketing issue; it is a decision-governance issue with customer trust consequences.

Practical implication: constrain inferred-state use cases and review the data inputs that can trigger sensitive offer changes.

ROI measurement and decision auditability

The article repeatedly links loyalty maturity to proving financial return rather than relying on engagement metrics. That shifts attention to measurement design, not just campaign performance. If loyalty decisions cannot be tied to monetary outcomes, they become hard to govern because teams cannot distinguish effective personalization from expensive noise. Auditability matters here because finance, marketing, and risk teams need the same evidence trail when programmes influence spend, retention, and customer allocation.

Practical implication: build an evidence chain from segment selection to spend outcome before scaling any loyalty decision engine.

NHI Mgmt Group analysis

Behavioral loyalty programmes are becoming decision systems, not just marketing layers. Once segmentation drives differentiated offers, the control question changes from campaign quality to decision governance. That means the real subject is not loyalty copy or promotion volume, but whether the underlying customer logic is explainable, bounded, and reviewable.

AI empathy creates a new governance burden because intent is inferred, not declared. The article shows brands moving toward signals that interpret life changes or emotional context. That is useful commercially, but it also means the programme is making judgments about customer state from behavioural proxies, which demands stricter oversight of what the model is allowed to infer.

Precision micro-segmentation sharpens the need for data provenance and access discipline. The more localized and granular the offer logic becomes, the more dependence there is on trustworthy customer attributes, clean source systems, and controlled edits by regional teams. The practitioner challenge is not personalisation at scale in the abstract, but preventing segment sprawl and uncontrolled decision drift.

Customer decisioning now carries audit expectations that used to sit only in risk and finance functions. The article’s ROI framing shows that loyalty teams must prove monetary impact, not just engagement lift. That makes auditability a programme requirement, because no high-value decision system should be allowed to influence revenue without evidence of why it behaved as it did.

Behavioral loyalty and AI micro-segmentation: this is the right concept for the next phase of customer governance. It describes a system where small behaviour changes, local context, and model-driven offers combine to shape customer treatment, and practitioners need controls that can explain and constrain that combination.

What this signals

Behavioral loyalty becomes a governance problem once segment logic starts changing who gets what and why. Teams should expect stronger demand for explainability, approval workflows, and reviewable decision inputs as customer personalization gets more precise.

AI empathy raises the bar for customer-data discipline. Once a programme infers mood, need, or life change from behaviour, the quality of the underlying signals matters as much as the offer itself. That makes data provenance and rule boundaries central to programme trust.

For practitioners

• Define segment governance rules Document who can create, modify, and retire customer segments, and require a review trail for every rule that changes offer eligibility or treatment.
• Validate customer-state inference inputs Review which behavioural signals can trigger empathy-based offers, then block inputs that are too weak to support reliable decisions.
• Tie loyalty logic to measurable outcomes Connect each major loyalty decision path to a measurable financial or retention outcome so finance and programme owners can test whether the logic is working.
• Control local override authority Limit how far regional teams can diverge from centrally approved segmentation rules, and log every exception that changes customer treatment.

Key takeaways

• The article shows loyalty moving away from generic discounting and toward behavioral psychology, AI empathy, and fine-grained segmentation.
• The governance pressure rises because more customer treatment decisions now depend on explainable data, local overrides, and measurable financial outcomes.
• Practitioners should treat loyalty decisioning as a controlled system, with documented segment logic, constrained inference, and auditable performance evidence.

Key terms

• Behavioural Segmentation: Behavioural segmentation is the practice of grouping customers by observed actions such as purchases, visits, or responses instead of by static demographics. In governed environments, those groupings can influence pricing, eligibility, and treatment, so the rules and data sources behind them need clear ownership and auditability.
• AI Empathy: AI empathy is the use of model-driven signals to adapt offers, tone, or timing based on a customer’s inferred state. It is not emotional understanding in the human sense. The governance issue is whether the inference is reliable enough to justify differentiated treatment.
• Decision Auditability: Decision auditability is the ability to prove why an access decision was made by linking the subject, action, resource, outcome and policy version. In modern identity programmes, it is the difference between reconstructing access after the fact and demonstrating control at the point of enforcement.
• Micro-segmentation: Micro-segmentation divides cloud environments into smaller trust zones so that identities cannot move freely between systems. It is a containment strategy rather than a prevention strategy. For NHI governance, it limits lateral movement and reduces the damage caused by compromised machine credentials.

Deepen your knowledge

NHI governance, agentic AI identity, and machine identity lifecycle are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are responsible for identity security strategy or NHI governance in your organisation, it is worth exploring.