Arcadia

Arcadia is a proprietary customer research platform designed to identify key moments in the customer journey. This tool enables strategists to segment and cluster audiences in real time, offering detailed data visualizations that help craft impactful and meaningful experiences. Arcadia was developed by a product team at Havas New York, where my role involved building various types of data-rich graphs using D3.js.

Survey design view showing psychographic questions for a Revlon drugstore makeup survey, with a multiple-choice question being edited inline

Edit journey modal listing journey phases — Awareness, Education, Trial, Purchase, and Advocacy — with the Trial phase actively being renamed

Analyze journey view filtering open-text responses by the keyword “Makeup,” with overlapping sentiment circles by age group and a list of respondent statements scored by sentiment

Keyword sentiment view comparing how age groups feel about “Makeup,” “Good quality,” “Cosmetics,” and “Drug stores” on a negative-to-positive scale

Average sentiment line graph plotting Under 35 and Over 35 respondents across the Awareness, Consideration, Purchase, Usage, and Advocacy stages, with a tooltip pinned to a question about cosmetics in drug stores

Sentiment journey graph with the filter panel open, segmenting respondents by annual income brackets from under $25,000 to over $200,000

Syndicated data comparison view showing horizontal bar charts that benchmark six Revlon survey versions across fashion attitudes, men's clothing brands, and foreign vacation spend

Analyze audience view with stacked bar charts breaking down lipstick purchase recency and relationship status by annual income segment

Strategy dashboard with a grid of saved insight cards — journey maps, stacked bars, heat maps, and violin plots — grouped by audience segment, with a hovered card revealing user notes

Insight export modal displaying a saved segmentation titled “Segmentation to further the overall value proposition,” with its question, stacked bar chart, and segment metadata ready to export