Case Study
Liberal Party of Canada and the 2021 Leaders' Debates
There are few inflection points in a campaign that are as critical as debates. Millions of Canadians watch live, millions more see coverage, clips, and commentary. For many voters, the debate represents the first time they’ve ‘dialed-in’ to the campaign. Underdogs can emerge, leaders can be toppled, and in one evening, campaign plans can go out the window.
Having an understanding of what happened during a debate, rooted in public opinion is critically important. Campaigns need to understand how their voters reacted so they can plan the days ahead. Will the leader be pivoting from a poor debate performance? Doubling-down on a competitor’s mistakes? Or striding away from a tie?
The Problem
In elections past, campaigns have relied on in-person focus groups paired with “dials” - a vintage technology that has participants turn their dial to the right when they see something in the debate that they like, and left when they dislike a moment in the debate.
Of course, in a pandemic election, this approach was not an option. More than that, this methodology is high-cost, high-risk, regionally narrow, and is a very atypical viewing experience.
Our Solution
We built a custom solution for the Liberal Party of Canada that enabled rapid feedback before, during and after the debate. This was a multi-modal approach, making full use of our recruitment and outreach methods. It included digital-first recruitment, SMS-texting, and virtual focus groups. This methodology was employed both in French and in English for their respective debates.
-
Participant recruitment - The hard truth of a campaign is that not every voter matters to everyone. Parties set their target, know who they can count on and who they need to persuade. We recruited voters in strategic ridings who were considering voting Liberal but had not yet ruled out other parties. Participants were recruited via targeted digital ads, paired with a comprehensive SMS-based vetting process, which allowed us to understand participants’ backgrounds and potential voting behaviour.
-
Pre-debate survey - Immediately prior to the debate, we had participants answer a series of questions via text, which created our baseline understanding. They were asked to evaluate each party leader, who they thought would win the debate, and how they currently planned to vote.
-
Debate interaction - We engaged with participants via SMS-texting, allowing them to remain focused on the debate while providing critical feedback. This interaction occurred in two ways - participants texted a 👍 when they liked what they were seeing from their assigned leader, and a 👎 when they disliked the leader’s performance. Participants were then asked who they thought won each section, and what the most memorable moment was. As input was generated, we used our proprietary data science to aggregate input and provide near-immediate summary data and visualizations to our client.
-
Post-debate survey - After the debate, we revisited the questions from the pre-debate survey, allowing us to understand how the overall opinion of the leaders shifted and how/if those impressions caused a change in vote intention.
-
Post-debate micro focus groups - A subset of research participants were asked to participate in a brief 20-minute focus group after the debate. This allowed us to have a more in-depth conversation about what stood out throughout the debate, and to drill down on the significance of specific moments where we observed an uptick in positive or negative sentiment.
Impact
This input was critical to the Liberal Party’s post-debate strategy development. It informed which moments to highlight during media interviews and in the creation of digital content. It also signalled what to look for in the following nights’ polling and analysis.