Quantitative data that captures how people feel about everything
Affie is an emotive thesaurus that allows users to search a word (e.g. trust) and see other concepts that are perceived to be emotionally similar (e.g. teacher, home, work boots). The tool provides quantitative data-driven insight into people’s emotional perceptions of anything; from a hospital room to a politician.
Research Question
How can consumer sentiment be used to inform creative processes and decision making?
Project Outcome
- The development of Affie, an online thesaurus for consumer sentiments.
- Identified a target market segment of creative professionals to use Affie within their workflow.
Affie uses an algorithm to show concepts with emotional (rather than literal) similarity.
Concepts within the database contain an emotional value based on three variable ratings: goodness, powerfulness and activity.
procedure
The project involved determining the potential market value of Affie for informing decision making within creative processes.
An early prototype of Affie was used for preliminary testing with over 100 potential users.
Participation was limited to individuals who work primarily in a creative process such as writing, designing, painting, advertising, event planning, marketing or broadcasting.
Participants were tasked to use the Affie prototype within one of their projects and complete a feedback survey at the end of the project.
Key insights
Awareness of the data-driven nature of Affie was essential for users to accept the legitimacy of search results.
People with limited experience in their creative process struggled to adopt the abstract results that Affie provided.
People with several years of experience in their creative process were far more capable of incorporating Affie into their workflow.
direction
Based on user research insights, two development paths were proposed:
Re-frame Affie as a tool for professional practices and expand the data set to allow for greater search result accuracy and demographic filtering of data.
Re-frame Affie as an easily accessible tool to find emotionally similar search results, vastly broadening the segment of potential users and usage scenarios.
development
Affie’s core product offering was re-framed as a general search tool algorithm that provides users with the ability to find emotional (rather than literal) search results. The data set was also expanded to provide emotionally similar words, images, colours and typefaces within search results to improve user accessibility.
Mock-up search result pages were used to run quick user testing of the product offering and validate market opportunity.
project funding
Affie was accepted into CSIRO's ON Prime program, a national science and technology accelerator that helps research team start ups discover real world applications for their projects.
involvement
I have been involved in the ongoing user research and testing of Affie prototype iterations, developing mock ups and interview discussion guides for the research team.