Gateway-Endpoint
Many doors, few rooms. Online radicalisation usually begins in communities and discussion spaces built around everyday interests, such as fitness, money, food and parenting. From there, a small share of the audience is walked one step at a time toward a hard political position. Gateway-Endpoint, a framework developed by Ripple Research, maps where people enter, where they end up and how they are moved between the two.
A cooking channel can end in distrust of the state without changing its look
On German-language TikTok, a family-lifestyle account with 31,000 followers posts about food, home remedies and family life. Its most-viewed video, with 1.7 million views, sets the food of earlier generations against breakfast cereal. Later videos question vaccinations before kindergarten. One ends with the creator saying she no longer trusts the system, filmed in the same warm kitchen light as the cereal video.
This is the femosphere. Young men meet the same pattern in the manosphere, where creators who start with gym routines and crypto tips can end on immigration or feminism. Gateway-Endpoint maps both.
The framework separates the doors people enter from the rooms a few reach
Gateways are the doors. They are everyday topics, from gaming and fitness to recipes and parenting, that draw a broad audience and build trust in a creator. Gateway content raises a grievance or offers help. It makes no political demand.
The hallway is the sequence of moves that carries part of an audience toward a political position. It has four steps: validate the frustration, blame a culprit, belong to an in-group, and act together.
Endpoints are the rooms. In the application shown here, they are the three core positions of the radical right. Nativism treats immigrants as a threat to a native group. Authoritarianism and hierarchy reject equality and call for strong rule. Conspiratorial anti-system belief holds that elites coordinate to harm ordinary people.
One rule does most of the sorting. If it raises a grievance, it is a gateway; if it prescribes political change, it is an endpoint.
Most of the reach stays at the doors
We built and tested the framework in a two-part study of the manosphere in the United Kingdom, Germany and France, covering TikTok, Instagram, YouTube and X. We coded 90 influencers and 60 political accounts, with 105.5 million engagements between them. In all three countries, the best-performing content stayed at gateway level or blended a gateway with a political claim, and purely ideological content reached fewer people. Discovery happened on TikTok and Instagram, the longer narrative on YouTube, and community on X and messaging apps.
Our study of the femosphere in Germany found the same structure aimed at women. In the lifestyle content we sampled, soft words such as love, family and natural outnumbered hard markers such as threat and danger by 35 to 1. Recipes, remedies, parenting and wellness replace gyms and crypto, and the rooms stay the same.
European surveys show a similar gap between the many who pass through the doors and the few who reach the rooms.
| Where | Who | Met hostile or hateful content online | Holds hard views | Source |
|---|---|---|---|---|
| EU | Internet users aged 16 to 29 | 49% (2023) | Not measured | Eurostat, 2024 |
| UK | 16 to 24 year olds | 54% | 3% advocate hateful views | HOPE not hate, 2024 |
| Germany | Adults | Not measured | 8% hold a clear far-right worldview (2023) | FES Mitte study, 2023 |
| France | Aged 15 and over | Not measured | 17% endorse hostile sexism | Haut Conseil à l'égalité, 2026 |
The surveys measure different things in different populations, so they do not add up to one pipeline. Only the UK figures come from a single sample.
The framework shows how content is built and how audiences move through it. It cannot show that any one viewer was persuaded, because neither of our studies tracked individual viewers over time, and the femosphere findings come from one platform in one country.
Gateway-Endpoint codes both ends of the path in one dataset
Research on online radicalisation tends to study one end of the path. Political scientists define the destinations: Cas Mudde's Populist Radical Right Parties in Europe (2007) and Pippa Norris and Ronald Inglehart's Cultural Backlash (2019) set out the positions our rooms are built on. Platform studies trace the route, as in Auditing Radicalization Pathways on YouTube (2020) by Manoel Horta Ribeiro and colleagues, which found users moving from milder to more extreme channels over time. Work on the manosphere, including Debbie Ging's Alphas, Betas, and Incels (2019), explains why the doors draw young men in.
We designed the categories and coding rules of Gateway-Endpoint to code both ends in the same dataset, post by post. That makes the hallway measurable, and comparable across audiences.
Use it to find the door where you can still reach people
A response works best at the door, before anyone has made a political claim. We map the doors for a given issue and audience and identify the creators who hold them, so organisations can speak on the same topics in the same tone. Our Meet-and-Move framework sets out how that response is built.
Platforms and researchers can use the same categories to code content over time. A wellness channel drifting toward vaccine refusal may never break a rule in any single post. Coding the sequence of posts is how the drift becomes visible.
Want to apply this framework to your issue?
Most engagements start with a conversation about what you are trying to understand, and where your audience is most open to a different story.