Postmedia Sports Betting Content

Finding opportunities within the newly legalized sports betting content space

Summary

Context

Organization: Postmedia Network Inc.

Industry: Sports news media, single-game sports betting

My role: UX Researcher

Team: UX team lead

Stakeholders: UX, Product, Monetization, Engagement, Marketing, and Editorial

Objectives:

To understand users who partake in sports betting and consume sports betting content
To gather feedback regarding our own sports betting content

Strategy

Research Type(s): Generative, evaluative

Methodologies: Competitive analysis, surveys, user interviews, contextual inquiry, thematic analysis, affinity mapping, rainbow spreadsheets

Timeline: 1.5 years

Participants: 812 survey respondents, 14 external users from the United States

Key Outcomes

Discovery research revealed five behavioural archetypes: as well as their journeys, jobs to be done, and pain points
That said: post-launch evaluation contradicted initial Marketing research and proved it to be inapplicable to contemporary audiences
Sports betting content was subsequently deprioritized, and partnerships not renewed

The Problem

With the passing of bill C-218 in Spring 2021, provinces were free to open the single-game sports betting market: with Ontario implementing these legislative changes in March of 2022. As a result, Postmedia found itself in an opportune position to take advantage of this fresh consumer market.
- According to Marketing, there was evidence to suggest that the Toronto Sun user base would be particularly receptive to sports betting content after legalization of single-game sports betting in Canada.

Process Overview

The Postmedia gambling/sports betting project spanned from January of 2022 to May of 2023 (~1.5 years)
Throughout this period, UX Research was responsible for two primary tasks:

Discovery research: to understand users who partake in sports betting and consume sports betting content, and how we might cater to these needs.
- i.e. User archetypes, journeys, needs, pain points, etc.
Post-launch evaluation: to gather feedback regarding the resulting Postmedia sports betting page
- i.e. Ascertaining the proportion of Toronto Sun website visitors who were actually consuming this content, as well as identifying barriers and pain points

Discovery Research

Action

14 individuals were recruited via Respondent.io to participate in remote user interview sessions
- These participants were chosen from a screener respondent pool of 177
US sports bettors--or individuals interested in sports betting--were targeted
- US participants were chosen because of the country's longer history of legal, single-game sports betting
  - i.e. We were able to recruit participants with a wider variety of experience levels
Each individual 30-minute session consisted of two components:
A semi-structured interview, with a focus on demographics, general behavioural habits, motivations/jobs to be done, and the high-level user journey
A contextual inquiry, focusing on pain points, areas of delight, and more granular detail regarding consulted platforms, resources, and specific use cases
Participants were compensated $30 USD for their contributions and time
Session data was then primarily analyzed by via a rainbow spreadsheet

Fig.1. Rainbow sheet showing some high-level, journey-relevant behavioural observations I made during my sessions with the 14 participants.

Findings

Between the screener survey responses and participant interviews, five behavioural archetypes were observed, each with distinct motivations for—and approaches to—sports betting
- e.g. While Sharps were primarily focused on monetary gain potential, Casual Dabblers were more interested in the social aspects of participating in sports betting.

Fig.2. The five behavioural archetypes that were identified in this study.

Altogether, these early behavioural archetypes documented insights regarding these users' jobs to be done, as well as goals, thoughts, pain points, feelings, actions, and needs (i.e. all of the makings of an empathy map)
In addition, early user journeys were constructed: including those for individual archetypes, as well as an aggregate to provide an overview
These were then to be validated post-launch

Gambling Discovery Research Insights - User Journey.png

Fig.3. High-level "cheat sheet" of the aggregated user journey for sports bettors.

Post-launch Evaluation

Action

UX played a minimal role in the design initiative for sports betting content after the initial archetype and journey research
Given that third-party vendors were providing content and assets, Product deemed UX's inclusion to be unnecessary
- That is: until there was concern regarding data from Google Analytics
The primary business function of Postmedia sports betting content was to drive traffic to affiliated sportsbook (i.e. sports betting platform) partners; that said, after launch, it was found that significantly fewer users were completing affiliate sign up flows.
- i.e. Postmedia sports betting content was underperforming
Results from a usability test regarding Scores and Stats pages presented similar insights
To understand why, data was gathered from a series of survey/study replications
- i.e. Done independently of the marketing team
The initial survey garnered 54 responses, being circulated via Mailchimp to members of the Toronto Sun user feedback panel
- i.e. Toronto Sun users who had volunteered to provide feedback in research initiatives

Due to their concerning nature, I conducted a replication to validate the findings of this initial survey, it was then repeated with different recruitment tactics:
- From March 8th-15th, the survey was appended directly to the end of editorial articles with the following call-to-action:
  “We are looking for insight and feedback regarding our Toronto Sun readers and the content they want to see. Please take a few moments to complete our short online survey."
- These stories were also shared on social media
- An additional call-to-action was included in the Toronto Sun's Midday Sun newsletter
- In total: 708 responses were garnered in the replication study.

Fig.4. Portion of early survey script, documented on Confluence

Fig.5. Article published in the Toronto Sun calling for survey participants.

Findings

In the case of both survey initiatives, the proportion of Toronto Sun visitors who have partaken in sports betting since its

legalization was significantly smaller than expected:
- The initial results showed that only 1/41 qualified respondents had done so
- Upon replication: 5.7% qualified respondents have placed a bet since legalization
The majority of the sports bettors who responded to this survey had at least 5 years of experience
Of the 510 respondents who answered this question, 67.1% were unaware that the Toronto Sun had sports betting content
Most importantly: most respondents simply had no interest in sports betting content
- In responses to the question "What were your thoughts on the sports betting content you saw on the Toronto Sun website?", the theme of disinterest was expressed 6.6x more frequently than any other feedback theme
  - The second most popular theme was outright disapproval of sports betting content

Fig.6. Survey response data (taken from Typeform report) showing that the vast majority of respondents had not participated in sports betting since its legalization in Canada.

Fig.7. Themes present in user responses to the question "What were your thoughts on the sports betting content that you saw on the Toronto Sun website?" Note that this was a follow-up to the question, "Do you recall seeing any sports betting content on the Toronto Sun website today?"

“Betting is addictive and I feel it has no place being promoted in a newspaper, tv, radio or internet.”

- Toronto Sun user response,

replication survey

Outcomes

Actions

In short, the initial findings of the Marketing department were found to be inapplicable to our current audiences
Given these results—along with the converging findings of several related projects completed across functional groups: sports betting content was subsequently deprioritized
- Likewise, the partnership with our then Scores and Stats provider was not renewed
Nonetheless, all research artifacts and findings were documented on Confluence, and organized into a project repository

Lessons Learned

Ultimately, I believe this project was a lesson in voicing my concerns more effectively, and insisting on due diligence
- In the kick-off meeting, I suggested that I conduct an additional cursory survey to gather user attitudinal data regarding sports betting and its place in the Toronto Sun
- Though I did insist, I ultimately acquiesced when I was told it would not be necessary
- Had I been more assertive in communicating my concerns regarding the data I was initially presented, as well as the importance of this additional survey, there might have been a possibility to avoid unnecessary costs
On another note, I have also changed my practice with regards to archetype- and persona-construction
- As can be seen on this very page: the archetypes I constructed had images associated with them
- These images, as well as some other traits documented in the full archetypes, were constructed as aggregated representations of the archetypal users
  - While they do literally give a face to these archetypes, they also run the risk of pigeonholing and depicting an over-simplified, one-dimensional mental model of their users
    - As such, I have stopped including irrelevant demographic traits and face images in these artifacts