Our foodfacts.org profile methodology

1. What our narrative analyses are, and are not

In addition to individual fact-checks, foodfacts.org publishes narrative-based analyses, including our “Misinformation Watchlist” profiles of individuals and organisations. These pieces examine patterns in how nutrition and food-system misinformation is framed and may be repeated over time, rather than only assessing single claims.

These analyses are not intended as personal attacks or as judgements on an individual’s character. Their purpose is to understand whether fact-checked claims sit within a broader pattern of messaging, and how those narratives may affect public understanding and behaviour.

2. How pages are selected

Individuals and organisations are only added to the watchlist or profiled if at least one of the following applies:

• Foodfacts.org has already published one or more fact-checks of claims they have made that were found to be false, mostly false, or misleading. This means we have been in touch with those individuals as part of our fact-checking policies and have given them an opportunity to comment on our reviews.
• They have been identified in reputable research reports or investigations as spreaders of misinformation about food, nutrition or related issues, for example in independent analyses such as this Changing Markets report and similar public-interest studies.

People are not listed or profiled solely because of disagreement with their opinions or dietary preferences. Inclusion is based on documented patterns in public content that are relevant to evidence-based nutrition and food-system debates.

Narrative analyses then get completed as and when resources allow. We prioritise the profiles who have the widest social media reach. Content from profiles with a smaller number of followers does not get reviewed for full narrative analyses.

3. Scope of content reviewed

Foodfacts.org applies a two-stage content review process for each full narrative analysis, designed to identify recurring patterns reliably before drawing conclusions about an individual’s or organisation’s messaging.

Stage one – identifying the range of themes. Before a profile is drafted, researchers review a minimum of approximately 9–12 pieces of public content from the individual or organisation, in chronological sequence. This threshold is informed by published methodological research on data saturation, which found that reviewing roughly 9–12 units of qualitative data was generally sufficient to identify the recurring range of themes present in a dataset, with few or no new themes emerging beyond that point (Hennink, Kaiser & Marconi, 2017). We apply this finding as a minimum content-review threshold for identifying an individual’s core, recurring claim patterns, rather than treating any single piece of content as representative on its own. Review continues beyond this minimum where new patterns are still emerging; the same published research found that a stable, representative set of themes is usually established once additional content stops introducing genuinely new patterns.

Stage two – understanding specific themes in depth. Once a recurring theme relevant to foodfacts.org’s remit has been identified, researchers may review further content specifically relating to that theme. The same research base underpinning our stage-one threshold shows that fully understanding the nuance of a theme typically requires reviewing a wider set of content than simply identifying that the theme exists (Hennink, Kaiser & Marconi, 2017). This deeper review may draw on content published outside the initial review window, including claims foodfacts.org has previously fact-checked, since these often provide the clearest documented examples of a specific theme.

In practice, this means:

• All content previously fact-checked that is associated with the individual or organisation is reviewed as part of this process.
• Publicly available posts and materials from approximately the three months preceding publication are reviewed as the primary basis for identifying current patterns. This may include relevant social-media posts, videos, blogs or public statements on nutrition, health, food systems or the environment.
• Where a specific theme warrants deeper analysis, review may extend further back in time to include earlier, previously fact-checked material relevant to that theme.

The analysis is limited to content that is already public and directly related to the topics covered by foodfacts.org. It does not extend to private life or material unrelated to food, nutrition or health narratives.

4. Analytical framework (F-A-C-T)

To ensure consistency and fairness, the same structured framework has been applied to all narrative analyses written since March 2026, which constitute the vast majority of our reviews. We are currently planning to update the analyses written beforehand. For each selected profile, the analysis considers the following dimensions:

F – Financial incentives. This examines whether there are visible revenue streams linked to the individual or organisation’s nutrition-related content, such as product sales, paid partnerships, subscription programmes or other commercial interests. Where such links are found to be relevant, they are described and sourced so that readers can assess potential conflicts of interest themselves.

A – Authority. This assesses how the individual or organisation presents, or is likely to be perceived as, an authority on the topics discussed. It includes professional credentials, training and experience, and how expertise is framed in biographies, titles or public materials.

C – Claims. This identifies the core recurring messages in the content, with particular attention to claims already assessed by foodfacts.org, other reputable fact-checkers or experts in science communication. Reviews consider what is emphasised, what is omitted, and how the messaging aligns or conflicts with the broader body of scientific evidence and public-health guidance on a given topic.

T – Tone (and framing). This examines how narratives are presented, including language that may amplify risk, dismiss mainstream guidance, promote overly simple solutions to complex issues, or frame evidence in polarised or sensational ways.

5. Evidence, sourcing and replicability

Each narrative analysis is based on specific pieces of public content that readers can examine themselves. Wherever practically and legally possible, foodfacts.org:

• Links to posts, articles or videos that are relevant to the analysis, including both previously fact-checked material and other representative content from the review period. Where a recurring pattern is illustrated by multiple similar examples, a single representative example is typically cited publicly for clarity and readability, rather than linking to every piece of content reviewed.
• References research reports or external analyses used to identify profiles.

The aim is to make the analysis as transparent as possible, so readers can see the evidential basis for any of our researchers’ own conclusions about recurring patterns or narratives.

In addition to the public-facing citations above, foodfacts.org retains an internal record of the initial pieces of content reviewed for each profile, namely the pieces that informed the review process described in Section 3. This internal record is separate from what is published, and exists to document that the stated review methodology was actually followed for a given profile.

6. Relationship with fact-checks

Narrative analyses are clearly distinguished from individual fact-checks on foodfacts.org. Because they focus on broader patterns and narratives, they necessarily rely more on the structured analysis and interpretation of our researchers than on a single verdict on a specific claim.

Where relevant, narrative analyses reference the underlying fact-checks and evidence base on which they rely. They are published as a complementary format to help readers understand how repeated claims may form part of a wider narrative, which may have more of an impact on their understanding of nutrition matters and behaviour than single claims.

7. Purpose and safeguards

The purpose of these analyses is to illuminate how repeated exposure to certain narratives can shape public perceptions of food, nutrition and health over time, even when individual claims might appear lower-impact when viewed in isolation.

To support fairness and accuracy:

• The same F-A-C-T framework is applied across profiles (from March 2026).
• The same two-stage content review process, described in Section 3, is applied across profiles.
• The analysis focuses on publicly available content and its potential impact on public understanding, rather than personal commentary.
• Profiles are intended to identify patterns in messaging and narrative, not to make claims about a person’s motives.
• Readers who believe a profile contains a factual error or omits important information can contact foodfacts.org through its corrections channels, and requests will be assessed under the site’s corrections policy.

Reference

Hennink, M. M., Kaiser, B. N., & Marconi, V. C. (2017). Code Saturation Versus Meaning Saturation: How Many Interviews Are Enough? Qualitative Health Research, 27(4), 591–608. https://doi.org/10.1177/1049732316665344