This Content Is Only For Subscribers
Ministry of Business, Innovation and Employment (MBIE) has released the latest Tourism Volumes & Flows data.
Domestic visitor counts are down 20 to 30 percent year-on-year, the days-per-visitor ratio is up by a third, and Queenstown and Wānaka are seeing some big shifts.
But the cause isn’t a change in how people are travelling, it’s a rebuild of the data itself. April 1, 2026, MBIE switched to a new telco data feed, and that switch is now working its way through the numbers.
Mike Webster, lead domain analyst (Tourism) at MBIE says, “Tourism Volumes & Flows moved to an improved telco data feed with a larger and more representative device panel, improved device identification, more accurate home-location assignment, improved treatment of international visitors, and better coverage of low-count areas. The updated feed makes use of the telco’s current network architecture and reduces exposure to legacy network changes, including the nationwide 3G shutdown.”
He says, “Data for periods before April 1 2026 has not been revised. The break is not intended to be permanent for all measures. MBIE and Vistr are investigating whether some historical components, including monthly unique visitor counts, should be revised, but will wait for a longer time series on the new feed before making that decision. Until otherwise advised, users should treat April 1 2026 as a change point in the series.”
The Domestic visitor drop
The question is whether that drop reflects real behaviour or simply how the new feed counts people.
Mike says, “The step change April 1 2026 is a measurement effect, not a change in visitor behaviour. Domestic unique visitor counts had been declining gradually before April, in the order of 5 percent a year. The additional step from April reflects the methodology change, which corrected an overstatement of domestic visitors in the previous feed. Visitor days are not affected in the same way, which is why the days-per-visitor ratio rises. The methodology notice sets this out in detail.”
He says “There is no formal margin of error published, although the user guide provides some further details on the limitations within the reported numbers. The updated outputs have been validated against independent benchmarks, including Stats NZ population estimates, the Domestic Visitor Survey, airport arrivals and transaction data.”
Regions to watch
Queenstown, Wānaka, Gisborne and Coastal Bay of Plenty still require a closer look before trusting the numbers. Mike says, “All outputs were comprehensively benchmarked against available indicators as part of implementing the methodology update. This highlighted that some regions were more significantly affected by the limitations of the previous feed, particularly areas with lower cellular coverage, smaller visitor counts, or a high proportion of repeat international visitation. Movements in these regions therefore reflect a combination of genuine market change and methodology improvement occurring at the same time.”
He says, “Given the time of year, this most affects Queenstown and Wānaka, which receive high levels of repeat visitation during the winter ski season. As set out in the methodology notice, the post-April outputs for these regions are supported by airport arrivals and other reference data and are considered more representative than outputs under the prior methodology. The large year-on-year increases reported are a combination of genuine growth and under-counting in previous years, and the caveat remains in place while corrections to historical outputs are considered.”
He says, “Gisborne and Coastal Bay of Plenty experienced significant weather events over the period of the change, making them more difficult to benchmark. Caveats will be reviewed once a longer run of data is available to make a complete assessment. There is no committed timeframe, but MBIE expects this review to take place over the next three to six months.”
Tracking visitor origin
It’s a fair question too. How the data can pin down where someone has travelled from without identifying them personally.
Mike says, “A visitor’s origin country is determined from the SIM card’s country of registration, together with the device’s usage patterns outside New Zealand before and after it appears in the country. Official travel statistics, including border-crossing data, are used only in aggregate to calibrate and validate the estimates. No person-level information is used, and no attempt is made to associate any individual traveller or border crossing with a specific device.”
The bottom line is, comparisons within the post-April period, or within the pre-April period, can be trusted. Comparisons covering from April 1 2026 especially in Queenstown, Wānaka, Gisborne and Coastal Bay of Plenty should be treated as indicative only until MBIE completes its review over the coming months.
