New Zealand insurers have relied on a familiar toolkit for monitoring customer experience: post-interaction surveys, formal complaints channels, and managers manually spot-checking a fraction of monthly calls. A deployment by Vero Insurance, announced on July 16, 2026, moves one insurer to a materially different model – and the timing aligns with the most specific regulatory pressure the sector has faced on exactly these functions.
The Financial Markets Authority’s (FMA) Financial Conduct Report for 2026/27, released on June 30, 2026, identifies the use of complaints data to drive improvements as a cross-sector priority, with insurers specifically named alongside banks, financial advisers, and investment managers. The FMA’s stated concern for insurers is precise: from its monitoring of the sector in 2025/26, only some insurers have comprehensive complaints policies that are regularly reviewed and updated, and overall, insurers are not yet consistently utilising complaints data to inform product changes or improve complaints processes. The regulator has signalled this work will continue into 2026/27.
Vero has implemented Salesforce’s Sentiment Analysis technology across its sales and service teams, becoming, according to the company, the first financial services organisation in New Zealand to use the tool. The system evaluates customer phone and email interactions in real time, classifying sentiment as positive, neutral, or negative. Where negative sentiment is detected, consultants receive a prompt to review the interaction and take action where required. The tool also flags potential complaints and indicators of customer vulnerability.
The technology was piloted between November 2025 and March 2026 across approximately 45,000 customer interactions before being extended across Vero’s consumer and business operations. Since launch, it has generated more than 65,000 sentiment outcomes, with approximately 10% of customer email and voice interactions now analysed in real time. Compared with Vero’s previous quality assurance approach, the system gives managers visibility across approximately 58 customer interactions per consultant each month – nearly 30 times the previous volume.
Nic Dorward (pictured), executive manager of consumer operations at Vero, said the technology changes when intervention is possible. “Insurance often comes at a stressful time in someone’s life. If we can better understand how a customer is feeling while we’re still working with them, we have a much greater opportunity to step in, resolve concerns, and provide the right support before those issues escalate,” she said.
The Conduct of Financial Institutions (CoFI) regime came into full effect on March 31, 2025, requiring licensed insurers to establish and maintain fair conduct programmes and comply with the fair conduct principle in all dealings with consumers. Since then, the FMA has progressively sharpened what it expects to see in practice. In April 2025, the FMA published its Fair Conduct Programme Insights Report, based on reviews of 38 fair conduct programmes (FCPs) – including those from six life insurers and 19 non-life insurers. The report found that while most FCPs were relevant and proportionate, consideration of evolving risks such as AI and digitalisation was limited across the sector. Critically, the FMA flagged that it intends to undertake a more in-depth review of FCPs with findings supported by evidence of how the programmes have functioned in action – a direct signal that documentation quality will give way to operational evidence as the primary test.
The FMA’s September 2025 complaints guidance is explicit about what good practice looks like: providers should use lead indicators to detect and prevent issues, not only respond to complaints once raised. The 2026/27 Financial Conduct Report reinforces that complaints are a lag indicator, and that the FMA expects insurers to complement complaints data with other lead and lag indicators. Real-time sentiment monitoring – detecting negative interactions before a formal complaint is lodged – is precisely the kind of lead indicator the regulator has described.
In a July 2025 report on insurers’ claims processes following the Auckland Anniversary Weekend flooding and Cyclone Gabrielle, the FMA also identified identifying and treating customers in vulnerable circumstances as an area requiring improvement. Vero’s system flags potential vulnerability indicators within interactions, giving consultants the ability to route customers to additional support services before an issue escalates.
Vero’s parent company, Suncorp Group, has been building AI capability across its Trans-Tasman operations over several years. In customer-facing functions, supported by platforms including Genesys and Salesforce, Suncorp uses AI for capabilities such as call summarisation, real-time agent assistance, virtual agents, automated quality assurance and sentiment analysis. The Vero sentiment monitoring deployment sits within that broader group technology strategy, which means it reflects coordinated platform investment across the Suncorp portfolio rather than a standalone initiative by a single brand. Suncorp also continues to build out VeroEdge, its intermediated business insurance platform, with additional broker partners expected to connect in FY26. For brokers working with Vero, the sentiment tool has a direct application: Vero says the capability can surface issues affecting shared customers earlier, reducing the likelihood of unresolved concerns becoming formal complaints.
The FMA’s 2026/27 Financial Conduct Report sets out four cross-sector themes, one of which is the use of complaints data to drive improvements – applying to insurers, banks, financial advisers, and investment managers. For insurers specifically, the FMA has noted that customer access to internal and external complaints processes is inconsistent, and that the sector needs to move toward using complaints insights holistically to inform product and service changes, not merely to resolve individual cases.
The traditional QA approach – sampling a small percentage of interactions each month – produces a narrow picture of what is occurring across a customer base. The 30-times increase in interaction coverage produced by Vero’s sentiment tool represents a structural difference in what management can see and, by extension, what the FMA will be able to assess when it examines whether a fair conduct programme is functioning in practice. Dorward framed the technology as reinforcing rather than replacing human judgement. “Customer advocacy starts with listening and acting with care. Sentiment Analysis helps our teams recognise when customers need us most – strengthening how we show up with empathy and intent,” she said.
The FMA has been clear that it views fair conduct as an ongoing operational commitment, not a documentation exercise. For insurers still relying on post-interaction surveys and manual QA sampling as their primary means of detecting complaints and vulnerability, the FMA’s stated intention to review how fair conduct programmes function in practice – and its specific 2026/27 focus on how insurers use complaints data – narrows the window for that approach to be considered sufficient.
