User research methods series: The Microsoft Reaction Card Method
Microsoft Product Reaction Cards can help researchers quickly understand people's gut reactions to products, designs or concepts in an objective and reliable way .
An example of Microsoft Product Reaction Cards being used in an online survey.
Originally developed at Microsoft in the early 2000s, the method involves showing people a product and asking them to pick out words from a set of pre-printed cards that describe their feelings towards it. The cards contain a wide range of positive, negative and neutral words like "accessible", "boring", "innovative", "frustrating", and so on.
In a typical study, participants are asked to pick 5 cards from a list of around 30 cards that represent their first impressions after a short interaction with the product. They then discuss why they selected each word card, allowing the researcher to capture rich qualitative data about specific likes, dislikes and emotional responses to what they have just seen or been using. Grouping and counting up the distribution of words chosen provides a simple way to quantify sentiment across multiple participants.
When used in the right way, the Microsoft Product Reaction Card method is a quick yet reliable method to gather feedback on the visual design of a product or service without being subjective or simply asking people if they like or dislike the design. The words act as prompts, lowering barriers for participants to openly discuss what they really think and feel, rather than what they think the researcher wants to hear. They can also be used to gather feedback on multiple designs to aid your decision-making. For best results, we recommend using an online survey tool to collect your data and use a sample size of at least 30 participants for meaningful results.
The method is lightweight, flexible and can be adapted for different use cases such as exploring early stage concepts or evaluating existing designs before a . It can work as an opening or closing activity before doing usability testing but it works best when treated as a separate study to limit bias and maximise the value a larger sample will provide.
