# Insight Memo Template Fill this in by 4:10 PM. Print or save as a one-page PDF for your portfolio. --- # What Yelp reviewers are telling Starbucks corporate **Researcher**: _______________________ **Date**: _______________________ **Data**: ~2,000 Yelp reviews of Starbucks locations across 10 metros, May 2026 snapshot of the Yelp Open Dataset. Reviews thematically coded with a starter codebook of 10 themes, extended by the researcher; every theme assignment grounded in a verbatim quote from the source review. --- ## TL;DR One sentence per insight. Three bullets max. The VP reads this in 20 seconds. - **Insight 1**: _____________________________________________________________ - **Insight 2**: _____________________________________________________________ - **Insight 3**: _____________________________________________________________ --- ## Insight 1. [Headline that names a business implication] **What we see in the data** 1-2 sentences. Cite a number (frequency, percentage, top-N). **Representative quotes** (verbatim, with stars + city) - *"_____________________________________________"* (1 star, Tampa FL) - *"_____________________________________________"* (2 stars, Philadelphia PA) **So what for Starbucks corporate** 1-2 sentences. Which exec owns this? What's the next move you'd recommend, or the next research question you'd ask? --- ## Insight 2. [Headline that names a business implication] **What we see in the data** **Representative quotes** - *"_____________________________________________"* (X stars, City State) - *"_____________________________________________"* (X stars, City State) **So what for Starbucks corporate** --- ## Insight 3. [Headline that names a business implication] **What we see in the data** **Representative quotes** - *"_____________________________________________"* (X stars, City State) - *"_____________________________________________"* (X stars, City State) **So what for Starbucks corporate** --- ## What this analysis cannot tell us (limitations) 3 bullets. The VP respects researchers who name their own limits. - Yelp reviewers are a self-selected slice of customers — disproportionately people with strong opinions, especially negative ones. These insights describe the people who chose to write, not the average customer. - _____________________________________________________________ - _____________________________________________________________ --- ## What I'd do next with another week 1-3 bullets. Where would you take this? - _____________________________________________________________ - _____________________________________________________________ - _____________________________________________________________ --- ## Method (for the appendix) - **Codebook**: starter codebook of 10 themes (mobile app, speed, drink quality, food quality, barista service, cleanliness, atmosphere/workspace, pricing, rewards, drive-thru), extended with [your added themes here]. - **Coding**: each review tagged with up to 3 themes by Claude Sonnet, with a verbatim quote required as evidence for every theme. Any theme whose supporting quote did not appear verbatim in the original review was discarded. - **Charts**: built in Vite + React + recharts; dashboard available at `localhost:5173`. --- ## Portfolio note If you're submitting this for a UXR portfolio, attach: 1. This memo as a 1-page PDF 2. A screenshot of your dashboard with one filter applied 3. The repo link with your extended codebook visible 4. (Optional) a 60-second Loom of you using the dashboard to answer one exec question