It has been estimated that around 80% of consumer decisions are driven by emotions.
Consumer Home and Personal Care giant, Unilever, was interested in using emotion analytics, an advanced form of text analysis, to understand more about the relationship between specific hair problems, good or bad hair days, and the emotional state of people using their hair products.
Unilever asked us to use emotion analytics, part of our Bear Upstream approach, to undertake a project with two specific phases:
Phase 1: Memory writing – When consumers talk about having a good or bad hair day, how does that make them feel, and do specific hair conditions such as frizzy or damaged hair, affect these feelings?
Phase 2: Product testing of two prototype hair products – Open-ended responses were collected from consumers following prototype trials. This phase used emotion analysis to explore differences between consumer groups and their feedback on two prototypes.
Understand what a good/bad hair day means to consumers and how specific hair problems can influence how they feel.
Explore two hypothesis that:
Participants were drawn from the Unilever consumer panel and were self-allocated to one of six hair-type categories. At the initial stage of analysis Hitch were not privy to the specific hair types and instead knew participants were only members of Groups 1- 6.
Phase 1 – Unilever has previously used a qualitative data technique called ‘memory writing’ (Haug,1992), which produces rich sources of qualitative data that lend themselves to in-depth analysis. Hitch designed a set of qualitative questions to allow participants to explore their memory for feelings around the concepts of good and bad hair days. Participants were asked, over a one week period, to complete a diary reviewing memories in the third person.
Phase 2 – Participants tested each prototype product for a one week period, changing products after the first week. Hitch again designed qualitative based questions that allowed participants to provide detailed text capturing emotions and feelings towards each product in terms of experience of the product in use, styling, post styling daily use and perceptions of feelings generated by use of the product.
All data from the diaries and product testing responses was collected via Unilever’s online system, and then uploaded into the emotion analytics software. Analysis is still underway but preliminary results show some very interesting findings.