Back to all articles
Blog post
September 21, 2026

How do different LLMs represent your organisation?

A customer evaluating a brand, a journalist researching a CEO, and a policymaker looking up a public agency are all doing the same thing — discovering organisations through AI. The difference is that they may each receive a different description, supporting evidence, and overall impression.

This shift represents a major change in modern communications.

Large language models (LLMs) now interpret information for audiences, rather than just helping them find it. Instead of listing search results, they retrieve content, select authoritative sources, and generate a single synthesised answer that many users accept without reviewing the original articles.

For PR, communications, and marketing professionals, this presents a new reputation challenge. Your organisation now has multiple AI-generated reputations, each shaped by the model your stakeholders use. Understanding these differences is becoming as important as understanding media coverage – in fact, the two are often closely linked.

There isn’t one AI version of your organisation

A common misconception is that ChatGPT, Gemini, Claude, and Perplexity all access the same information. In reality, they  don’t. 

At a high level, all leading LLMs are built on similar foundations. They’re trained on vast collections of books, websites, news articles, public documents and licensed datasets that help them understand language and generate human-like responses. Increasingly, they’re also capable of retrieving live web information, allowing answers to incorporate recent events rather than relying solely on historical training data.

However, their similarities end there.

Each LLM uses a unique combination of training data, retrieval architecture, and ranking logic. As a result, each model answers three key questions differently before generating a response:

  • What information should I retrieve?
  • Which sources should I trust most?
  • What deserves emphasis in the final answer?

These decisions fundamentally shape how organisations are represented. To explain this better, we summarise how different AI models consult different publications and finally cite them in AI answers:

LLMs are not impartial, but come with their own weights and balances. In a sense, it reflects dynamics we already see at play across PR & Comms. 

If four experienced journalists had to write a profile of the same CEO after attending the same press conference and have access to the same reports, one may write about leadership, another  financial performance, another governance and another might frame the story around innovation. None are necessarily wrong—they’re simply viewing the event through a lens of individual expertise, and therefore, making different editorial decisions.

LLMs behave in similar ways. As a result, organisations are now represented by multiple AI-generated narratives, rather than a single authoritative digital narrative.

The same prompt can produce four different narratives

These differences are most apparent when users ask AI questions that require judgment rather than simple factual recall.

Consider a prompt like: “Which are the leading banks in Southeast Asia?” Across ChatGPT, Gemini, Claude, and Perplexity, you will likely see many of the same names — DBS, UOB, OCBC, and Maybank. This overlap occurs because all four models recognise these institutions as major regional banks.

However, the models differ in their explanations of why these banks are considered leaders.

ChatGPT may highlight DBS’s digital banking leadership and customer experience. Claude is potentially more likely to discuss governance, regional strategy, and long-term institutional strength. Gemini may emphasise recent awards and publicly available web information, while Perplexity often presents answers as research comparisons supported by multiple citations. These changes are not necessarily so big as to be immediately noticeable, but over time the results accumulate. 

An organisation may not meaningfully change between two users asking two different LLMs , but the reputation narrative shifts accordingly. This distinction matters because stakeholders rarely ask AI for isolated facts. They ask questions like:

  • Should I work with this company?
  • Which university is most innovative?
  • Has this government agency delivered on its commitments?
  • Who are the market leaders in this sector?

AI responds by interpreting credibility, authority, and context, rather than simply retrieving documents.

For communications teams, this means your organisation is increasingly evaluated through comparative prompts, where competitors, industry peers, and institutional benchmarks appear alongside your organisation by default.

Every LLM has its own citation fingerprint

When ChatGPT, Gemini, Claude, or Perplexity answer a question, they first retrieve documents from distinct information ecosystems. A recent study analysed 17.2 million AI citations across ChatGPT, Gemini, Claude and Perplexity and found that each model retrieves from substantially different source ecosystems rather than a shared pool of webpages. Two LLMs can answer the same question accurately while relying on entirely different publications.

Instead of viewing these as technical differences, it is more useful to consider them as citation behaviours. Each model consistently references different types of publications when explaining organisations. 

  • ChatGPT – Building consensus from multiple sources

ChatGPT functions more like an executive briefing writer than a traditional search engine.

Rather than listing multiple links, ChatGPT typically combines information from several credible publications into a coherent narrative. If mainstream media, company information, and industry commentary consistently describe an organisation as a market leader, ChatGPT is likely to reinforce that positioning, regardless of the original source.

For a prompt like “Tell me about Singapore Airlines”, a typical ChatGPT response integrates its history, customer experience, awards it has won over time, its financial performance, etc., into a cohesive description. Individual articles become almost invisible, as the model prioritises a coherent narrative in its answer over transparently showcasing all citations. 

For brands, this means consistency across publications is very valuable. ChatGPT values credible signals that repeat across multiple publications over one off news moments.

  • Gemini – Reading the living web

Gemini approaches organisations differently, as its retrieval is closely linked to Google’s broader information ecosystem.

When we ask  “What has Enterprise Singapore done to support AI businesses?”, Gemini is more likely to include recent programme announcements, official government webpages, and newer online reporting. Its responses often feel more current because they draw from a web ecosystem designed to reflect continuously updated information.

For government agencies, this has practical implications. Official announcements and well-structured public information become machine-readable assets that help AI explain policy more accurately. These become important information touchpoints into how audiences understand government through AI.

  • Claude — explains the ‘why’, in addition to the ‘what’

Claude’s defining feature is its emphasis on context.

Other models often prioritise concise answers but Claude frequently elaborates on why an organisation is respected. Questions about leadership, governance, ethics, and institutional reputation tend to produce more detailed explanations rather than brief summaries.

For a prompt like “Why is DBS considered one of Asia’s leading banks?”, Claude is more likely to discuss how the bank has regionally expanded, its digital transformation, what’s unique about its leadership etc. Its responses resemble analyst reports more than search summaries, as it favours high-authority editorial and institutional content.

For executive communications, this makes Claude particularly influential. Leadership narratives and corporate values often receive more contextual treatment than with other models.

  • Perplexity — makes your media strategy visible

Perplexity changes the experience by treating citations as a central feature rather than a supporting detail.

A question comparing two sustainability leaders, for example, typically returns numerous linked sources from business media and research publications. Users can immediately review the origin of each claim.

For PR teams, this creates greater transparency. Publication quality becomes visible within the user experience and is not hidden behind an AI summary. This means media strategy influences the evidence presented alongside the perception of the organisation. 

How commercial brands and government agencies are represented differently

Although LLMs behave similarly across sectors, the questions users ask differ fundamentally.

  • For brands, AI compares before customers get the chance 

Consumers rarely ask AI what a brand does. They definitely do ask which brand is better.

Questions like:

1. Which airline has the best customer experience?
2. Is Salesforce better than HubSpot?
3. Which bank is most innovative?
4. Who are the leaders in cloud computing?

These questions are inherently comparative.

As a result, your organisation is often introduced alongside competitors before stakeholders visit your website.

During a recent Isentia webinar on Measuring your brand’s visibility in AI answers, we analysed airlines across multiple LLMs with Lumina AI View. Using identical prompts, different models highlighted different airlines and emphasised varying strengths, such as premium service, operational performance, innovation, and customer experience. Competitive positioning shifted depending on the model, even though the underlying organisations remained unchanged.

For commercial communicators, competitor association is becoming as important as share of voice. AI evaluates brands in context, not in isolation.

  • AI becomes the interpreter of policy for government agencies

Public sector organisations encounter a different reputational challenge. Audiences increasingly ask questions that begin with how, why and can I trust:

1. What support does this agency provide?
2. Has this ministry achieved its policy goals?
3. What AI initiatives has this department introduced?
4. Has this programme faced criticism?

LLMs then synthesise official publications and institutional information into a single accessible explanation. This changes the role of public communications. The objective is not to ensure AI models have sufficient credible, authoritative context to explain complex policy accurately.

As AI becomes the primary interpreter of government information, communications teams are responsible for managing both visibility and understanding.

Measuring AI reputation with Lumina AI View

Reputation has been shaped across three familiar environments: media, search and social. AI introduces a fourth environment.

LLMs condense dozens of publications into a single response, reducing the traditional discovery process between a question and an opinion.

During the recent webinar, Prashant Saxena, VP of Revenue and Insights says, “We used to Google it. Now we’re ChatGPT-ing it.”

This behavioural shift is significant. Reuters Institute research found that only a small proportion of AI chatbot users regularly click through to original articles. 

Practically, every piece of earned coverage now has two audiences:

  • The people reading it.
  • The AI models learning from it.

Both shape reputation.

This is the challenge Lumina AI View was designed to solve.

Instead of measuring whether an organisation appears in ChatGPT, AI View measures how different LLMs represent the organisation. Using consistent question sets across ChatGPT, Gemini, Claude, Perplexity, Grok, and Google AI Overviews, it compares narratives, citations, and competitive positioning through a unified reputation methodology.

The framework is built around four familiar dimensions: Direction, Performance, Integrity, and Innovation. Rather than tracking keywords alone, AI View identifies which reputation pillars dominate AI responses, which publications influence those narratives, and how representation differs across models. It also gives the organisation an AI score based on all the above factors.

The tool also shows organisations which publications are covered by different AI models. If the organisation’s AI score on the tool is high for ChatGPT, but struggles with or is a lower score on Gemini or Perplexity, the tool allows you to see where you’re underperforming. If there are publications that are covering the wrong information, or the important ones just don’t show up, then that’s an indication to why the AI score is low on that platform. 

What this means for PR and communications professionals

The communications profession has always adapted to new discovery channels, from newspapers to search engines, and from social media to digital news.

AI is different because it does not simply help audiences find information, but it increasingly determines how organisations are introduced.

For communications leaders, this changes media strategy in three key ways. 

  • First, publication quality becomes more important than quantity, as different LLMs repeatedly reference authoritative sources when constructing narratives around your organisation.
  • Second, visibility for the organisation’s spokespeople now extends beyond interviews and articles. The expertise attributed to leaders increasingly shapes how AI explains an organisation’s direction and credibility. 
  • Finally, competitor positioning is now continuous rather than campaign-based, as AI naturally compares organisations whenever users ask for recommendations or leadership insights.

Organisations that succeed in this new environment are those that have leadership, performance, integrity, and innovation represented consistently across every major AI model, regardless of where audiences and stakeholders begin their search.

Want to see which sources are shaping how AI describes your organisation? Get in touch about Lumina AI View.

Share

Similar articles

object(WP_Post)#8084 (24) { ["ID"]=> int(50662) ["post_author"]=> string(2) "75" ["post_date"]=> string(19) "2026-11-04 01:49:28" ["post_date_gmt"]=> string(19) "2026-11-04 01:49:28" ["post_content"]=> string(21250) "

How do I navigate aged care in Australia? How easy are government services to deal with? How well supported are people with disabilities? These are big questions about life in Australia, and many people now put them to an AI tool before they visit a website, pick up the phone or speak to anyone. They rarely mention a government agency by name, yet the answer they get still shapes what they believe about the services behind it: whether the system works, who is responsible and whose information they can trust. That is reputation taking shape, in a conversation no communications team is part of.

To see how that plays out, we used Your Prompts, a new feature in Isentia's AI View, to track how leading AI models answer these questions. Here's what we found, whose information the models used, how each model goes about sourcing its answers, and what it all means for public-sector communicators.

Starting with the question

Most AI visibility tracking starts with an organisation's name. Your Prompts lets you set your own questions, which gives public-sector teams a much wider view of their reputation. People rarely ask an AI tool about an agency by name; they ask for help with whatever problem is in front of them, whether that's getting a parent into aged care or working out what support is available for a carer. Or they ask something bigger, such as how well the country looks after older people.

We tracked six sets of questions across leading AI models, covering the things people deal with at different stages of life: finding their way around government services, getting older, living with disability and caring for someone, raising a family, staying healthy, and making ends meet. For this article we focused on three of them: access and navigation, aged care, and disability and carers. These are the moments when people depend most directly on the government, which means a wrong or incomplete answer can have real consequences for someone trying to get help, and can colour how they feel about the system as a whole.

We compared answers from ChatGPT, Claude, Gemini, Perplexity and Google AI Overview, looking at what each model said, which sources it cited and which organisations it named. The questions were set up for tracking rather than drawn from a survey of what the public actually asks, and the findings reflect a single point in time, using the latest data from 27 October 2026. Even so, the patterns in how AI finds and presents information about public services were clear and consistent.

How easy are government services to deal with in Australia?

We asked this alongside a second question: how well government support reaches the people who need it. Most models landed on a similar view: digital services work well for simple tasks, while complex cases, phone contact and people in vulnerable situations are where things get harder. The more interesting part is where that view comes from. The models formed it largely through oversight bodies, with the national audit office's findings carrying particular weight.

Government's own voice only came through when a model cited official pages, as when one model described digital services as a major strength, using a government website as the source. In other words, when AI assesses how the government is performing, it largely takes the word of the ‘referees’. Owned content can earn you a line in the answer, but auditors, commissions and advocates write most of the scorecard. For reputation, that gives an audit finding a second life. Well after the news cycle moves on, it can still be part of the answer people get when they ask how the government is doing. That brings us back to a question we raised in our article on AI visibility for government agencies: is your organisation still seen as the main authority on its own services?

How do I navigate aged care in Australia?

Almost every model named government services as the place to go for financial assessments, payments and in-person help, and the tone was neutral to positive. One model went as far as recommending that people complete a means assessment early so they don't end up paying more than they need to.

The catch is that the agencies people were being sent to were rarely the source of that information, which mostly came from other government portals and guides written by aged care providers and advocacy groups. In one answer, even an agency's own contact details were taken from a seniors' advocacy group's website. Being named in an answer, then, is quite different from being its source. It builds recognition, but someone else is setting people's expectations of what they'll find when they get there, and if that guide is out of date, it's the service they feel let down by.

Zoom out to the bigger question of how well Australia looks after older people, and the agencies that were the first port of call in the practical answers didn't appear at all. These answers were framed by the Royal Commission into Aged Care Quality and Safety, which concluded in 2021, with most models opening on its findings and some giving them more weight than the reforms that followed. This is the reputational point that's easiest to miss: the practical question tells people where to go, while the big question tells them what kind of system they're dealing with. As we covered in why content freshness matters for AI visibility, old stories have a habit of shaping AI answers long after the facts have moved on.

How well supported are people with disabilities in Australia?

The models discussed income support payments without naming the agency that delivers them, focusing instead on the NDIS and the gaps in support. That said, not every absence is a problem. Some questions genuinely aren't about a particular organisation, so it makes sense for it to sit them out. The useful distinction for communicators is between a relevant absence, where the question simply isn't about you, and a missed opportunity, where your work is being discussed and you aren't getting the credit. Repeated across enough answers, a missed opportunity leaves people with a picture of support that doesn't include the people delivering it.

Questions about caring for someone played out differently. Every model gave almost the same answer and pointed people to the same national carer support service. When official information is clear and consistent across agencies, peak bodies and support services, AI repeats it faithfully. There was one notable exception, where a model explained the government's carer payments using a blog from a private company that sells personal medical alarms. The information was mostly accurate, but it shows how far an explanation can travel from its original source. The good news is that consistency across your wider network really does pay off: when the organisations around you describe your services the way you do, AI is more likely to get them right, and that shared story becomes part of your reputation.

Why each model tells the story differently

Across all three categories, each model showed clear habits in the sources it reached for. We can see what each one cites, even if we can't see exactly how it decides, and the patterns are consistent enough to act on. In practice, your reputation can look different depending on the tool someone uses. These habits also build on what we found when we looked at how different LLMs represent your organisation.

ModelWhere it gets its informationExample from our researchWhat it means for communicators
ChatGPTOfficial statistics, audit reports and formal reviews. It pairs almost every claim with a figure or finding, much like a policy analyst writing a briefing note.When asked whether government support reaches people in need, it built its answer around the national audit office’s findings. It was also the model most likely to cite the website of the agency delivering a service.Oversight reports shape how ChatGPT judges your performance, while your own pages shape how it describes what you do.
ClaudeThe widest mix of sources, including advocacy groups, provider guides and commentary alongside government material. It tends to explain the reasons behind an issue and present more than one side.When describing the Royal Commission’s “national disgrace” finding, it relied mostly on second-hand accounts of the report, such as Wikipedia and a private aged care provider’s website.The organisations that retell your story effectively become your spokespeople.
GeminiUniversity research and international bodies such as the OECD, drawn from Google’s broad index of academic platforms, video and international sources.On aged care, it drew on an overseas policy institute’s summary of Australia’s new aged care legislation.Published research and evaluations carry real weight, as do secondary sources describing your work, wherever in the world they are published.
PerplexityThe most recently updated pages, quoted closely. It is built around live search with visible citations, so it favours recent, easy-to-access pages.On access to services, it quoted a government corporate plan published only weeks earlier.This is where keeping your own content current pays off fastest.
Google AI OverviewA small number of sources, summarised briefly, because it is designed to give a quick answer at the top of search results.On disability support, it relied mainly on the scheme’s own website and summarised criticism in general terms, without naming who made it.The few sources it chooses carry a lot of weight, so clear, well-structured official pages give you the best chance of being one of them.

One question, several versions

Put together, the same question can produce an audit-style answer, an advocacy-informed analysis, a research summary or a current snapshot, depending on which tool someone happens to open. Two people asking the same question can come away with quite different impressions of the same service. There's no single AI view of an organisation to manage, because each model has its own sources and each needs a slightly different approach.

Zooming out: how these answers add up to reputation

Look at any one of these answers and it can seem like a visibility issue. Put them together and it's a reputation issue. For public-sector organisations, reputation has always been about more than what's said about you by name. It rests on whether people believe the system works, know where to turn and trust what they're told, and those impressions are now forming in AI answers, one question at a time.

Across these questions, four things shaped the impression an answer left:

  • Who gets named. Whether you appear when your work is being discussed, and get the credit for it.
  • Who does the explaining. Whether people hear how your services work from you, or from a provider, an advocate or a company blog.
  • Who acts as referee. Whose assessment of your performance gets repeated as the verdict.
  • Which story sets the frame. Whether the answer opens on where things stand today, or on a finding from years ago.

None of this shows up if you only ask AI about your organisation by name. You see it when you ask the questions your audience is asking. And it matters, because AI is now one of the main places where people go for information about the things that matter to them.

What this means for public-sector communicators

Government organisations can't control what AI says, but they can do a lot to influence the information it draws on, and with it the reputation that forms around them. Five practical steps came out of this research.

  • Track the big questions, not just your name. The most useful insights came from questions about services and real-life situations rather than about any one organisation, so start with what your audience is trying to do or understand, and check what AI is telling them.
  • Work with the bodies that assess your performance. Auditors, commissions and advocates shape how AI judges your work, so respond to their findings publicly and make sure your progress is recorded in the sources they, and AI, rely on. That way an old finding doesn't have to stand alone.
  • Support the organisations that explain your services. Sister agencies, peak bodies, providers and advocacy groups often explain your services for you, so give them clear, current information to work from. As we found in our analysis of Tier 1 and Tier 2 media in AI answers, specialist and niche sources often matter more to AI than their profile suggests.
  • Keep practical information clear and up to date. Plain-language pages on eligibility, processes and how to get help are what AI repeats when people ask how to do something. Organisations that deliver services directly rely on this more than most, because outdated information can lead someone to miss out on support they're entitled to.
  • Check answers across several models regularly. Each model uses different sources, so one model's answer is a poor guide to the others, and because answers shift as new content is published, regular tracking shows whether your updates are reaching the sources AI uses.

Whose information is AI using?

AI is already answering big questions about life in Australia, whether or not the public-sector organisations involved are part of the conversation. This research shows that an organisation can be mentioned often and still have surprisingly little say in how its services are described or how its performance is judged, with other organisations doing much of the explaining and each AI model weighing those voices differently.

For communicators, the job is to understand that wider network of sources and make sure accurate, current information runs through all of it. It's worth asking what AI says about your organisation. It's just as worthwhile asking what AI says when nobody mentions you, and whose information it's using to say it. That's where much of your reputation is now taking shape.

Want to see what AI tells people about the questions that matter to your audience? Get in touch about AI View.

" ["post_title"]=> string(63) "How do LLMs represent public services to the public they serve?" ["post_excerpt"]=> string(150) "Learn how AI cites government organisations, whose sources shape its answers on public services, and what public-sector communicators can do about it." ["post_status"]=> string(7) "publish" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(4) "open" ["post_password"]=> string(0) "" ["post_name"]=> string(62) "how-do-llms-represent-public-services-to-the-public-they-serve" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2026-11-04 01:49:36" ["post_modified_gmt"]=> string(19) "2026-11-04 01:49:36" ["post_content_filtered"]=> string(0) "" ["post_parent"]=> int(0) ["guid"]=> string(32) "https://www.isentia.com/?p=50662" ["menu_order"]=> int(0) ["post_type"]=> string(4) "post" ["post_mime_type"]=> string(0) "" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" }
Blog
How do LLMs represent public services to the public they serve?

Learn how AI cites government organisations, whose sources shape its answers on public services, and what public-sector communicators can do about it.

object(WP_Post)#12195 (24) { ["ID"]=> int(50575) ["post_author"]=> string(2) "75" ["post_date"]=> string(19) "2026-10-23 01:20:42" ["post_date_gmt"]=> string(19) "2026-10-23 01:20:42" ["post_content"]=> string(35642) "

A weekly tracker tracing the stories, figures and policies driving media and social coverage in the 2026 New Zealand General Election across Online News, TV, Radio, Newspaper, Magazine, X, Facebook, Instagram, TikTok, Threads and Forums. This edition covers conversations and coverage from 15th - 21st October, highlighting what’s shaping the election conversation.

1. Election Conversation

10,572
total mentions this week
2,842 media7,730 social

Media = news coverage. Social = all public posts across X, Facebook, Instagram, TikTok and Threads - everyone posting, not just media outlets.

2. Media Issue of the Week

Opportunity Party Puts the Kingmaker Question at the Centre of the Campaign

Opportunity (TOP)271 media items · 325 social posts

The Opportunity Party, led by Qiulae Wong, was the most covered new story in the news this week. It polled 8% in the 1News Verian poll on 13 October and 5.4% in the Post/Freshwater poll on 19 October, just above the 5% needed to enter Parliament. With National and Labour close together, outlets described it as a possible kingmaker. Christopher Luxon ruled out working with the party at the Herald leaders' debate, then told West Coast voters he would work with "whoever I have to", and campaign chair Simeon Brown ruled it out again. David Seymour said ACT would consider a four-party arrangement that includes Opportunity. Later in the week the focus turned to the party itself, with former board members describing internal problems and Wong saying the party's rules need another look.

News outlets ran 271 items on the party or the kingmaker question in the headline or summary (9.5% of the 2,842 election items), and it was mentioned somewhere in 724 (25.5%). Coverage was highest on 15 October, after Luxon's West Coast comments, Brown's response and Seymour's remarks, and again on 20 and 21 October, with new poll results, the former board members' claims and Wong's comments on coalition deals.

It matters because it affects who can form a government. The Post projected that Opportunity's seven MPs could give either a Labour-led bloc a narrow 61-seat majority or a National-led bloc 66 seats, so every party has had to say whether it would work with Opportunity. On social media it was a smaller conversation, with 325 posts (4.2%), and those posts were less negative than average (44.4% negative against 53.0% overall). The story is about coalition options rather than a policy position, so it is not linked to a party issue below.

3. What's happening across media and social?

2,842 media items · 7,730 social items · share of each channel's own conversation, so the two bars are directly comparable

"Social" is all public posts across X, Facebook, Instagram, TikTok and Threads - not just media organisations' own accounts.

MediaSocial
Issues
Cost of Living
Media
13.9%
Social
18%
Law & Order
Media
9.1%
Social
22.1%
Economy & Jobs
Media
16.5%
Social
12.5%
Health
Media
11.8%
Social
11.9%
Infrastructure & Transport
Media
10.1%
Social
4.1%
Parties
National
Media
25.8%
Social
21%
Labour
Media
23.5%
Social
17.8%
NZ First
Media
12.7%
Social
19.6%
Greens
Media
13.7%
Social
18.2%
ACT
Media
9.9%
Social
14.1%

Most frequent bylines and presenters: The most frequent bylines and presenters this week were Sean Plunket (The Platform), Ryan Bridge (Three), Henry Cooke (Waikato Times) and John MacDonald (Newstalk ZB).

Widest-reach accounts: The accounts with the widest reach among election posts were news outlets: the New Zealand Herald, 1News, Radio Tarana, RNZ and Stuff, along with the New Zealand Labour Party account. Party leaders' own accounts posted often but have smaller followings.

Law and order made up 22.1% of the issues discussed on social media but 9.1% in the news, about 2.4 times as much. Part of the gap is the continuing reaction to the Greens' and Te Pāti Māori's visit to a Mongrel Mob event, last week's Media Issue of the Week: 137 of the 791 law and order posts mention it. Across all social posts the visit came up in 291 (3.8%), 70.5% of them negative, and in 257 news items (9.0%).

The Opportunity Party shows the gap the other way round. It was in the headline or summary of 9.5% of news items but only 4.2% of social posts, about 2.3 times as much attention in the news. Among the parties, NZ First, the Greens and ACT had a bigger share of social mentions than news mentions (NZ First 19.6% against 12.7%, Greens 18.2% against 13.7%, ACT 14.1% against 9.9%), while National and Labour had a bigger share in the news (National 25.8% against 21.0%, Labour 23.5% against 17.8%).

4. Party Share of Conversation

1. National23.2%
2. Labour20.4%
3. NZ First16.5%
4. Greens16.2%
5. ACT12.2%
6. Te Pāti Māori6.5%
7. Opportunity (TOP)5%

Shown for seven parties, including Te Pāti Māori and the Opportunity Party (TOP), which is this week's Media Issue of the Week. Shares add up to 100% of the seven shown. Share of conversation measures attention, not polling, and is not a sign of electoral success.

5. Top candidates this week

1. Christopher Luxon1,614
2. Chris Hipkins1,611
3. Winston Peters1,133
4. David Seymour840
5. Chloe Swarbrick412

Christopher Luxon (1,614 mentions) and Chris Hipkins (1,611) were level at the top. Both were at the centre of the final leaders' debate on 20 October and the questions about working with Opportunity, and each party announced new policy during the week.

Winston Peters (1,133) was third and was mentioned far more on social media (726) than in the news (407). The most shared posts about him covered his comments on diversity, his message to David Seymour on strategic voting and his public meetings. David Seymour (840) drew attention for saying ACT could work with Opportunity.

Chloe Swarbrick (412) was in the news more than on social media (227 against 185), with coverage of the Greens' plan to reverse public service cuts and the Mongrel Mob visit. Mentions measure how much someone is being talked about, not how they are polling.

6. Election Issues

Top 4 themes this week

Cost of Living1,621
Economy & Jobs1,605
Law & Order1,430
Health1,252

What's driving conversation about each party?

Each party's top issues this week in social conversation, ranked by how much they were discussed

National
1Health
2Jobs
3Cost of living
Labour
1Health
2Education
3Cost of living
Greens
1Cost of living
2Law and order
3Training wage
ACT
1GP access
2Cost of living

Issues are ranked from social posts. Counts are not shown because the numbers for a single party and issue are small.

7. Issues to Watch

  • The coalition question around the Opportunity Party is likely to stay in the news until election day. Luxon has given two different answers on working with the party, and polls this month have put it between 4.8% and 8%, so a small shift in support could change which parties can form a government. The party's internal problems are also now in the spotlight.
  • Voter enrolment is becoming a political issue ahead of the 25 October deadline. New rules mean people cannot enrol at the polling booth, and 112 news items this week covered enrolment or the electoral roll, many focused on young and first-time voters. The number enrolled when the roll closes will show whether these concerns made a difference.
  • Social conversation leaned negative this week: 53.0% of posts with a clear tone were negative, 32.6% neutral and 14.4% positive. Cost of living is a top issue for four parties (National, Labour, Greens and ACT). Health is a top issue for two parties (National and Labour). Issues that matter to several parties are the ones most likely to keep shaping the campaign.


If you want to find out more about how Isentia can help you understand the NZ media landscape, please reach out to us here.

" ["post_title"]=> string(97) "NZ Election 2026: TOP as Kingmaker, Law & Order Surge on Social, and Minor Parties Gaining Ground" ["post_excerpt"]=> string(152) "Analysis of the latest NZ Election 2026 media tracker: The Opportunity Party's kingmaker potential, Law & Order social trends, and minor party momentum." ["post_status"]=> string(7) "publish" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(4) "open" ["post_password"]=> string(0) "" ["post_name"]=> string(92) "nz-election-2026-top-as-kingmaker-law-order-surge-on-social-and-minor-parties-gaining-ground" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2026-10-23 01:20:50" ["post_modified_gmt"]=> string(19) "2026-10-23 01:20:50" ["post_content_filtered"]=> string(0) "" ["post_parent"]=> int(0) ["guid"]=> string(32) "https://www.isentia.com/?p=50575" ["menu_order"]=> int(0) ["post_type"]=> string(4) "post" ["post_mime_type"]=> string(0) "" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" }
Blog
NZ Election 2026: TOP as Kingmaker, Law & Order Surge on Social, and Minor Parties Gaining Ground

Analysis of the latest NZ Election 2026 media tracker: The Opportunity Party’s kingmaker potential, Law & Order social trends, and minor party momentum.

object(WP_Post)#7855 (24) { ["ID"]=> int(50464) ["post_author"]=> string(2) "75" ["post_date"]=> string(19) "2026-10-19 01:20:16" ["post_date_gmt"]=> string(19) "2026-10-19 01:20:16" ["post_content"]=> string(32656) "
A weekly tracker tracing the stories, figures and policies driving media and social coverage in the 2026 New Zealand General Election across Online News, TV, Radio, Newspaper, Magazine, X, Facebook, Instagram, TikTok, Threads and Forums. This edition covers conversations and coverage from 8th - 15th October, highlighting what’s shaping the election conversation.

1. Election Conversation

18,417
total mentions this week
4,479 media13,938 social

Media = news coverage. Social = all public posts across X, Facebook, Instagram, TikTok, Threads and Forums - everyone posting, not just media outlets.

2. Media Issue of the Week

Green and Te Pāti Māori MPs Draw Criticism Over Meeting With Mongrel Mob Members

Greens / Te Pāti Māori112 media items · 116 social posts

Candidates and sitting MPs from the Green Party and Te Pāti Māori attended a meeting with Mongrel Mob members at a gang address. When photographs from the meeting were shared on social media, the response from commentators and rival parties was swift and largely critical.

This was not the week's highest-volume story, accounting for only 112 of 4,479 media items (2.5%, after filtering out unrelated international content) and 116 of 13,938 unique social posts (0.8%, after collapsing retweet/repost duplicates down to one count per message), well below the volume generated by ongoing leaders' debate coverage, which totalled 1,125 media items. Two independent methods nonetheless support treating it as a genuine, separable story: an exact keyword count aligns closely with an unsupervised clustering analysis of the entire media population, which identified it as a distinct 47-item cluster - smaller than general campaign coverage and several specific policy announcements, but clearly distinguishable from them. What sets the story apart is the intensity of the reaction rather than its scale: 72% of social commentary on the topic was negative, the most one-sided sentiment recorded for any subject this week, and the average post attracted roughly 5.7 times the typical level of engagement. Taken together, these figures point to a genuine public reaction concentrated on social platforms, and the episode provides National and ACT with a law-and-order narrative that may continue into coming weeks. It has not been linked to a tracked party policy below, as it reflects a matter of conduct rather than a policy position.

3. What's happening across media and social?

4,479 media items · 13,938 social items · share of each channel's own conversation, so the two bars are directly comparable

"Social" is all public posts across X, Facebook, Instagram, TikTok, Threads and Forums - not just media organisations' own accounts.

MediaSocial
Issues
Cost of Living
Media
15.8%
Social
22.7%
Economy & Jobs
Media
19.4%
Social
17.5%
Health
Media
11.7%
Social
9.9%
Infrastructure & Transport
Media
11.4%
Social
5.2%
Law & Order
Media
5.3%
Social
10.6%
Parties
National
Media
30.9%
Social
27.5%
Labour
Media
25.3%
Social
21.6%
NZ First
Media
14.7%
Social
21.5%
Greens
Media
14.1%
Social
20.4%
ACT
Media
6.8%
Social
6.4%

The most active commentary this week came from Sean Plunket, the Chris Chang and Tova O'Brien pairing, and John MacDonald.

By follower reach, the most prominent social accounts posting about the election this week were major outlets' own accounts - the New Zealand Herald, 1News, RNZ, Stuff and Radio Tarana - alongside the official Labour Party account. Individual leaders' personal accounts posted actively but with smaller followings than these outlet and party accounts.

The Greens continue to receive disproportionately greater attention on social media relative to traditional news coverage, accounting for 20.4% of social party conversation compared with 14.1% in media coverage - roughly 1.4 times the media share.

Law and order shows a similarly pronounced gap between channels this week, accounting for 10.6% of social conversation compared with 5.3% in the news. This is consistent with the public reaction to the Mongrel Mob meeting, which appears to have generated considerably sharper engagement on social platforms than mainstream coverage volume alone would suggest.

4. Party Share of Conversation

1. National33.1%
2. Labour26.6%
3. NZ First20.6%
4. Greens19.7%

Shown for the four largest parliamentary parties (National, Labour, Greens, NZ First); ACT, Te Pāti Māori and TOP are tracked elsewhere in this report but not included in this chart, so these shares sum to 100% of the four shown, not of all tracked parties. Share of conversation measures attention and mentions only - it is not polling and should not be read as an indication of electoral success.

5. Top candidates this week

1. Winston Peters1,931
2. Christopher Luxon1,631
3. Chris Hipkins1,466
4. David Seymour916
5. Chloe Swarbrick561

Winston Peters leads the rankings this week, a shift driven primarily by a widely circulated moment in the minor-party leaders' debate in which he described diversity as "a weakness rather than a strength" while delivering a pointed critique of the left-leaning bloc. Christopher Luxon and Chris Hipkins follow closely behind as the campaign moves into its final weeks, while David Seymour and Chloe Swarbrick complete the top five, each drawing attention largely through their parties' own policy announcements this week.

6. Election Issues

Top 4 themes this week

Economy & Jobs3,014
Cost of Living2,901
Health1,789
Infrastructure & Transport1,508

What's driving conversation about each party?

Each major party's top-ranked issues this week, by real topic volume in their social conversation (after excluding other parties' names, individual people, and procedural or horse-race chatter), with how that conversation leaned

NegativeNeutralPositive
National
Cost of living · 86 items
Neg 46.5%Neu 31.4%Pos 22.1%
Tax reform · 34 items
Neg 44.1%Neu 52.9%Pos 2.9%
Fiscal plan · 42 items
Neg 38.1%Neu 57.1%Pos 4.8%
Labour
Cost of living · 88 items
Neg 30.7%Neu 36.4%Pos 33%
Fiscal plan · 47 items
Neg 55.3%Neu 34%Pos 10.6%
Economy · 70 items
Neg 34.3%Neu 48.6%Pos 17.1%
Greens
Wealth tax · 41 items
Neg 53.7%Neu 34.1%Pos 12.2%
Cost of living · 43 items
Neg 27.9%Neu 41.9%Pos 30.2%
Jobs · 112 items
Neg 34.8%Neu 33.9%Pos 31.2%
ACT
Fiscal plan · 24 items
Neg 50%Neu 37.5%Pos 12.5%

Volume and sentiment above are based on social media data only.

7. Issues to Watch

  • The meeting between Green and Te Pāti Māori representatives and Mongrel Mob members represents a reputational risk for both parties heading into the final weeks of the campaign. Whether National and ACT are able to sustain this as a law-and-order narrative through to the 7 November election is worth monitoring closely.
  • Labour's fiscal plan issue recorded the most negative social sentiment of any tracked policy this week, at 55.3% negative, followed by Greens' wealth tax at 53.7% and ACT's fiscal plan at 50.0%. All three touch on tax, spending or fiscal policy specifically, suggesting the fiscal debate is where social sentiment is running most negative across the board this week, rather than any one party facing an isolated problem.
  • Social media discussion of the Greens continues to centre on polling performance and tactical voting considerations rather than policy substance. Coalition and tactical-voting framing accounts for 12.7% of their social conversation this week, compared with just 2.2% focused specifically on their wealth tax proposal - a pattern that has now held for three consecutive weeks. Whether that balance shifts as the campaign enters its final stretch would be worth tracking.

If you want to find out more about how Isentia can help you understand the NZ media landscape, please reach out to us here.

" ["post_title"]=> string(83) "NZ Election 2026: Mongrel Mob Controversy, Fiscal Policy and Greens on social media" ["post_excerpt"]=> string(156) "Explore New Zealand’s 2026 election conversation, from the Mongrel Mob meeting controversy and law and order to reactions to Labour’s capital gains tax." ["post_status"]=> string(7) "publish" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(4) "open" ["post_password"]=> string(0) "" ["post_name"]=> string(49) "nz-election-2026-mongrel-mob-fiscal-policy-greens" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2026-10-19 01:20:22" ["post_modified_gmt"]=> string(19) "2026-10-19 01:20:22" ["post_content_filtered"]=> string(0) "" ["post_parent"]=> int(0) ["guid"]=> string(32) "https://www.isentia.com/?p=50464" ["menu_order"]=> int(0) ["post_type"]=> string(4) "post" ["post_mime_type"]=> string(0) "" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" }
Blog
NZ Election 2026: Mongrel Mob Controversy, Fiscal Policy and Greens on social media

Explore New Zealand’s 2026 election conversation, from the Mongrel Mob meeting controversy and law and order to reactions to Labour’s capital gains tax.

Ready to get started?

Get in touch or request a demo.

Contact us Request a demo