MIT has free course lectures online. Here's what they said about low-cost market research for entrepreneurs.
-"If you do not ask, you cannot get an answer. You need an explicit set of hypotheses you want to test in advance."
-Design your type of analysis in advance as this will determine sample size. Too-large sample size can make results seem significant when they are not, really. Sample size does not need to be huge.
-Think about actionability.
-Myths of market research: Market research is only for big decisions. Instead, relatively trivial decisions can be improved by a small amount of market research, and strategic mistakes can cost a lot in the long run - research is worth it and does not have to be expensive.
(Train clients to come to us more frequently?)
-Statistical analysis - can use Chi-Squared for categorical and for interval questions (like how much do you agree/how good do you think the food system is) can use 2-sample t-test or one-way ANOVA (for multiple groups, such as comparing segments), and MANOVA and MANCOVA for multiple differences between groups
-To test multiple changes at once (e.g. price and packaging) can use a fractional factorial design.
-Other analysis includes correlation, regression, logistic regression, discriminant analysis
No comments:
Post a Comment