1) Who we want to target. We want to work with the brand to understand the composition of their clientele. Do they target high-income or low-income users? Men or women? Young adults or middle-aged folk? Is it a professional line or a casual line? Conventional or avant garde? 

2) Sample size for this population. Once we figure out the group of individuals we are targeting, we will be able to calculate the accurate sample size at 95% confidence interval (alpha of .05). We will require this number of responses in order to determine statistical significance of our experiment.

3) We can gather the necessary data in a low-cost and high-effect manner by presenting our ten photos on an Instagram feed – called Testagram. We want to utilize the fact that people enjoy scrolling through and liking Instagrams on fashion bloggers’ feeds. 

4) We will reach the accurate target population by publicizing the Testagram through “influencers” that relate to our target clientele. We will reach out to these influencers and have them publicize the Testagram feed. As fashion brands have struggled to reach a youthful, representative demo in a decentralized marketplace, many have turned to influencers — internet celebrities whose market worth is determined by the size of and engagement with their online followings — as a means of spreading awareness.

5) Statistical Standardization: We propose creating an Instagram and comparing the likes received across the 10 photos. In our future statistical analysis, we will use “position” on the feed as a covariate, to account for the fact that users may get bored after 5 photos and leave the site. We will also standardize our results for each user – that way the likes from a user who likes all 10 photos will be weighted differently than the likes from a user who only likes 2 photos.