Sample: n=4,200, 18-75 years old.

Demographics: Nat rep, Great Britain sample. Quotas on Age, Gender, Region and Social Grade

Platform: Online

LOI: 20-25 mins

Fieldwork dates: 13 August – 5 September 2025

Categories: Seven categories included (Sportswear, Chocolate, Supermarkets, Toilet Paper, Personal Care, Quick Service Restaurants, Banks & Building Societies). Each respondent answered about brands in one of the categories.

How we measured cause commitment: 60 causes included overall. Each respondent answered on 20 causes. Each cause assigned randomly. To develop the list of causes we used the UN’s SDGs as a start point and then fine tuned the list ensuring we had a topical and relevant list for GB today. The question wording for cause commitment is as follows: There are many issues or causes discussed in our society today, covering a wide range of topics such as diversity, inclusion, fairness, community safety, health and wellbeing, education, the economy and environmental protection. In this section, we'd like to understand which, if any, of these causes or issues you feel personally connected to. (Commitment Scale 1-10)

We used a number of different types of analyses to dig deep into the data:

• Factor Analysis (a statistical method to identify underlying relationships amongst sets of variables. Enabled us to group our 60 causes into nine broad themes)
• Multi-Dimensional Scaling Mapping (a technique to map the similarities of various elements across many dimensions. Allowed us to explore the relationships between factors, brands, and to see where key subgroups map against these)
• Ipsos Bayes Nets (IBN - IBN is the gold standard key driver analysis used within Ipsos. Enabled us to determine drivers per category and the importance of ‘do good’ associations versus functional & emotional aspects)
• Impact of ‘doing good’ on Brand Closeness, Advocacy and Price Perceptions (we used several kinds of regression analyses to determine the magnitude of the effects driving brand measures).
  • For overall importance and strength of relationship we employed bivariate linear regression to assess the explanatory power (Adjusted R2) of associations with “doing good” towards brand KPIs overall and for each category.
  • For identifying uplifts from associations of “doing good” towards brand KPIs we employed a proprietary, respondent-level Random Forest Regression algorithm with partial dependence, controlling for overall affinity for each brand.