Are boards measuring the right things?

An organisation can make customer transactions faster, increase digital engagement and reduce service costs while leaving unanswered a significant commercial question: are those improvements strengthening the customer relationship?

KPMG New Zealand’s CX50 2026, part of a longstanding global customer-experience research programme, draws on 10,502 evaluations of 80 brands across eight sectors by 1,076 New Zealand consumers.

The research explores what KPMG calls ‘liquid expectations’ – the tendency for customers to judge organisations against the best experiences they encounter anywhere, regardless of industry. It reveals tensions between technology adoption and customer satisfaction, convenience and security, and perceived value and the overall customer experience.

For chairs, the findings raise questions about the evidence supporting the board’s understanding of organisational performance, particularly when evaluating strategic investments intended to improve efficiency, customer experience and long-term commercial returns.

Performance and importance tell different stories

KPMG assesses customer experience against six pillars: personalisation, time and effort, expectations, integrity, resolution and empathy.

Time and effort, which measures how easy organisations make interactions for customers, received the highest average performance score of 7.8 out of 10.

Yet KPMG’s modelling found that integrity had the strongest relationship with customer advocacy, with a weighting of 21.1%. Personalisation had the strongest relationship with continued customer loyalty, at 22.6%. Empathy received the lowest performance score, at 6.6 out of 10.

The percentages represent the relative weight of each pillar in KPMG’s statistical model, rather than the proportion of customers identifying a particular factor as important.

Organisations can therefore perform strongly on convenience without achieving equivalent strength in the characteristics associated with recommendation or continued customer preference.

Reduced transaction times, higher self-service uptake and lower contact-centre costs can demonstrate operational success. They do not necessarily establish whether customers trust the organisation more, regard its services as relevant or intend to remain with it.

Investment decisions require evidence that improvements in operational efficiency are delivering the customer outcomes assumed in the business case.

More AI, mixed results

Telecommunications recorded the highest reported AI usage among the sectors surveyed, at 21%, but the lowest satisfaction with those interactions. Only 56% were rated satisfactory, with an average satisfaction score of 6.4 out of 10.

Non-grocery retail recorded considerably lower AI usage, at 6%, but satisfaction reached 73%, with an average score of 7.3.

The comparison does not establish that greater AI adoption produces poorer customer experiences. The sectors differ in the services they provide, the complexity of customer enquiries and the circumstances in which assistance is required.

Across the study, virtual assistants were the lowest-rated customer service channel, scoring 6.72 out of 10. Apps performed best, at 8.05, followed by face-to-face interactions at 7.77, websites at 7.65 and email at 7.51.

AI assistants featured in 11% of the customer–brand relationships surveyed, accounting for 1,132 reported interactions over the previous two years. Of those interactions, 66% were rated satisfactory, 25% neutral and approximately 10% dissatisfied.

A programme may meet its deployment targets, process substantial transaction volumes and deliver cost reductions. Its contribution to customer satisfaction, retention and the wider commercial relationship requires separate evidence. 

Customers are reluctant to surrender control

The report also examines consumers’ openness to allowing AI agents to undertake activities on their behalf over the next 24 months.

While 37% were open to AI helping choose products or services, only 27% were open to delegating tasks to an AI agent owned by a trusted brand, compared with 49% who were closed to the idea. Openness declined sharply among consumers aged 55 and over.

In a separate assessment of acceptable AI roles, 44% considered it acceptable for AI to escalate sensitive or complex situations to a human, while only 30% accepted AI completing tasks automatically without their input.

KPMG describes AI agents as a potential ‘second front door’ to organisations, through which customers may search, compare, negotiate and transact on their behalf.

This introduces questions about identity, consent and authorisation, particularly when AI systems act independently. The findings suggest customers are more comfortable receiving assistance than handing over control.

For organisations investing in AI-enabled services, the research raises questions about how much autonomy customers will accept and whether adoption forecasts reflect those preferences.

Friction has a purpose

In a case study featured in the report, BNZ says customers can appreciate additional verification steps when they understand how those measures protect them from scams and fraud.

The bank has also migrated more than one million customers away from legacy physical authentication methods, simplifying access while strengthening security.

The example challenges the assumption that removing friction always improves customer experience. Additional steps can provide reassurance and strengthen trust, even when they make a transaction less convenient.

Value and experience are different measures

Perceived value and overall customer experience can produce different assessments of the same organisation.

Mitre 10 topped the customer-experience rankings with a score of 8.19, but its perceived-value score was lower at 7.67.

Kmart recorded the opposite pattern, with a customer-experience score of 7.52 and perceived value of 8.16.

The two retailers illustrate how a strong customer experience and a strong value proposition can produce different results. Neither measure establishes which organisation has the stronger commercial performance.

KPMG’s research does not examine board reporting practices or establish that directors are relying on inappropriate measures. Nor does it quantify the financial returns associated with the customer-experience characteristics it identifies.

An organisation may successfully deliver the operational improvements it promised while the evidence of stronger customer relationships remains less convincing. Board reporting needs to establish whether those improvements are delivering the commercial outcomes assumed when the investment was approved.

Read The big shift: experience by design – New Zealand CX50 2026 here. Contact Lance Webb, Head of Customer Experience Practice KPMG, for more information.