Skip to main content

What we still do not know

The evidence base is strong on Māori authority, relationships, ethics and the history of mapping. It is much thinner on what happens to Māori GIS systems five, ten or twenty years after they are built. This page keeps those gaps visible rather than filling them with confident assumptions.

Highest-priority gaps

GapWhy it mattersPriority
Five-, ten- and twenty-year survival of Māori GIS projectslaunch success says little about long-term usabilityvery high
Real hapū GIS total cost of ownershiparchitecture debates lack Māori-specific longitudinal financial evidencevery high
Administrator succession failurespractical control can disappear when one person leavesvery high
Consultant exit and account ownershipcontracts directly affect practical rangatiratangavery high
GIS-specific implementation of the Māori Data Governance Modelcurrent Māori governance guidance is broader than geospatial architecturevery high
Māori spatial-sensitivity taxonomypublic/private is often too crudehigh
Metadata leakage from Māori GIS portalstechnically plausible but little Māori-specific empirical evidence existshigh
Cloud-provider backup deletionterms vary and are difficult for small organisations to evaluatehigh
Māori GIS cyber incidentslittle public case literature is availablehigh
Long-term preservation of StoryMaps and web mapsinteractive applications are harder to preserve than datasetshigh
AI spatial inference over Māori cultural datacapability is advancing faster than governance evidencevery high
Costs of over-restrictionprotection is better studied than lost benefit from excessive restrictionmedium-high
Effects on rangatahi accessprotection can unintentionally reduce transmission and capabilitymedium-high
Longitudinal SaaS versus open-source experiencestrong opinions exist, comparative Māori evidence is weakvery high

We do not have a catalogue of failed Māori GIS systems

Public research provides numerous examples of successful or promising Māori mapping projects and broader evidence about capability and trust problems. It does not provide a defensible list of Māori organisations whose GIS systems failed after funding ended.

Do not invent one.

The absence of published cases does not mean failures never occur. It means the evidence is insufficient to name organisations responsibly.

Cost evidence is modelling, not observation

The Ten-year cost models are planning models based on realistic architecture and labour assumptions. They are not the results of following five hapū GIS programmes for a decade.

A valuable future research programme would document:

  • annual platform cost
  • volunteer and staff hours
  • consultant support
  • hardware replacement
  • migration events
  • outages
  • funding changes
  • user numbers
  • survival after staff turnover

We need more post-project research

Case studies often end when the new tool launches. Useful longitudinal questions are:

  • Is it still used?
  • Who administers it now?
  • Have data been migrated?
  • Is the original community still able to access the information?
  • Are licences still paid?
  • Did the project create internal capability?
  • Was knowledge returned from researchers and consultants?
  • Can the system survive its original champion leaving?

AI is changing the boundary of what can be inferred

There is strong evidence that AI can infer location and automate feature extraction. There is much less Māori-specific empirical work on how those capabilities affect wāhi tapu, cultural landscapes, historic maps, environmental data and other Māori spatial information.

This should remain an active research area rather than a page written once and forgotten.

Contributions and corrections

This resource should be revised when better evidence becomes available. Useful contributions include:

  • published evaluations
  • iwi or hapū public reports
  • peer-reviewed papers
  • public procurement and governance guidance
  • documented platform migrations
  • transparent cost cases
  • corrections to technical claims

Sensitive or confidential project failures should not be published merely to make the site look comprehensive.

Editorial rule

Where evidence ends, say so. Separate:

  • established fact
  • evidence-supported interpretation
  • practitioner experience
  • site recommendation
  • open research question

That distinction is part of the resource's data-governance practice, not just an editorial preference.

Last reviewed: 16 August 2026