Since the summer of 2024, MTD has maintained a Canadian media ecosystem seedlist to track political discourse. The dataset underwent an annual review and update in the fall of 2025, which also included an expansion into the US political influencer space1. Since its launch, we have used the data from the seedlist to research and report the state of the political discourse in Canada, culminating in reports such as the Meta News Ban Brief and the Canadian Influencer Brief. Notably, it played a crucial role during the 2025 Canadian Federal Election. We will end this guideline series by presenting a brief case study on how the seedlist was expanded and utilized for near real-time research during an active election cycle.
Work around the seedlist for the federal election began in February 2025, as we anticipated that a Canadian federal election would be called shortly. We began by recruiting data reviewers in preparation to add and update election candidates, as well as any new political influencers who may emerge during the election cycle. Once the writ of election was issued on March 23, 2025, we recorded seed data for each candidate using the official candidate list of each major political party (Liberal, Conservative, NDP, Green, Bloc Quebecois, and People’s Party). The team updated existing seeds if a candidate was already present in the seedlist as a politician or influencer and created new seeds for candidates who were new to the seedlist. This made up the majority of the seedlist work during the election.
At the same time, other teams at MTD were carrying out ongoing monitoring and analysis of issues that arose during the election. Data collected from Canadian seeds were used to assess if election issues flagged by our public tipline were gaining as much traction among the political elites in the seedlist. This assessment was repeated in the offline realm through our survey team. We also had a team of research assistants, called the “boots on the ground” team, who created social media avatars to monitor new topics around the election that were being recommended by their social media algorithm. The cooperation between the different teams created a positive reinforcing cycle. Emerging key influencers would also be flagged by the monitoring team to be added to the seedlist for our election monitoring efforts.
One of the issues flagged around the start of the advanced polling was the spread of election fraud claims. These were first flagged by both the boots team and the public tipline. Then we analyzed how widespread the fraud claims were using our collected social media data. Since the seedlist included all current candidates, we were able to measure if any candidates were discussing or even disseminating claims of fraud. Our analysis was able to show that most election fraud discussion came from influencers and only the People’s Party of Canada (PPC) candidates were disseminating these claims.
Beyond analyzing election fraud claims, the seedlist was crucial in our efforts to update and inform the public during and after the Canadian federal election in 2025. Our full report on the election presents the findings of our multi-disciplinary monitoring effort during the entire campaign period.
The 2025 federal election demonstrated what a mature seedlist makes possible: the ability to answer questions while they are still relevant. Because the infrastructure was already in place and data was already collected for the main seedlist, our team could focus on event-specific expansions of the seedlist that could facilitate analysis within a short period of time. This feature of the seedlist is applicable to many other important political events and research questions. We hope this guide helps other research teams build the same foundation.
- A special thank you to the team of research assistants during the 2024 and 2025 seedlist enrichment efforts: Ashley Vu, Olivia Melanson, Nader Hedfi, Kiran Gill, Zahra Mustin, Liam O’Toole, Bela Sullivan. ↩︎
Leave a Reply