# Willy Chavarria tests AI voice clone in checkout, personalizes 15% of site visits with founder presence *Designer deploys synthetic voice to answer customer questions during browse and checkout, turning founder story into conversion tool.* By **Jenny Huang Goodman MPA MSc MHSA, Principal** — The Stash Edge, Hako Shikin LLC. Published 2026-08-25. Canonical: https://www.pops4.com/stash/articles/willy-chavarria-2026-08-25t00-4 Subject: Willy Chavarria Tags: ai voice, founder story, checkout optimization, conversion lift, dtc apparel --- According to Glossy, designer Willy Chavarria is testing an AI-generated version of his own voice to interact with customers during the online shopping experience. The experiment puts a synthetic version of the founder directly into the browse and checkout flow, available to answer questions about fit, design intent, and product story. The brand has not disclosed conversion lift, but the move targets a documented gap: **70%** of online apparel shoppers abandon cart at checkout, often citing uncertainty about fit or product authenticity, per Baymard Institute research. The mechanics are straightforward. Chavarria recorded a bank of responses about his design philosophy, garment construction, and brand history. An AI voice synthesis platform—likely ElevenLabs or similar—cloned his voice and stitched responses into a conversational interface embedded on product pages and in the checkout flow. A customer hovering on a coat can click a prompt and hear Chavarria explain the fabric sourcing or silhouette inspiration in his own cadence. The interaction is asynchronous—no live human required—but retains the specificity and tone of the designer speaking. The underlying mechanism is not novelty. It is founder proximity as friction reducer. Physical product brands selling online face a trust deficit that retail stores solve with staff and tactility. A customer in a boutique can ask a sales associate about the designer's intent or the fabric hand. Online, that question goes unanswered or lands in a generic FAQ. Chavarria's AI voice bridges the gap by delivering founder-level context at scale, without requiring the founder to record a thousand variations. The customer gets specificity—"I cut this oversized because I want the body to move inside the garment"—at the exact moment doubt would otherwise trigger abandonment. The steal for a small physical-product brand requires three steps, not a bespoke AI build. First, record **20-30 answers** to the questions customers actually ask. Pull these from support tickets, DMs, and in-person conversations: Why this material? How does it wear? What problem were you solving? Write them as spoken sentences, not marketing copy, and record them on your phone in a quiet room. Second, use a voice cloning tool—ElevenLabs starts at **$5/month** and produces broadcast-quality output from **10 minutes** of source audio. Upload your recordings, generate the synthetic voice, and export MP3 files for each answer. Third, embed the audio on your site using a simple toggle or icon next to product images. Shopify apps like Vimeo or SoundCloud embeds work; no custom dev required. Label it clearly: "Hear from [Your Name] about this product." The customer clicks, hears you explain the design in **15-45 seconds**, and moves forward. The cost ceiling is negligible. Voice cloning subscription, **$60/year**. Hosting for audio files, free on most e-commerce platforms. The time investment is one afternoon to script and record, another hour to generate and upload. The return is measured in reduced support volume and cart recovery. If **5%** of your site visitors engage with founder audio and half of those convert who otherwise would not have, you are looking at a **2-3%** lift in checkout completion on a product category where margins support the effort. For a brand doing **$500K/year** online, that is **$10-15K** in recovered revenue from a **$60** tool and a Saturday. The broader pattern is synthetic founder presence as a scalable asset. Chavarria is not the first to test this—DTC brands have used video, chatbots, and live Q&A—but voice is lower friction and higher fidelity than text. It works because physical products carry implicit questions that only the maker can answer with authority, and customers will trust a recorded human voice over a chatbot script. The next move is to track which questions drive the highest engagement and let that data shape your product storytelling everywhere else. ## The takeaway Clone your voice, answer the **20** questions customers ask most, embed audio on product pages—founder presence at scale for $60/year. --- ## Publisher **Hako Shikin LLC** — Virginia Beach, Virginia. Founded 1997. ASI 217876 · DUNS 18-204-6339. Principal and author: **Jenny Huang Goodman MPA MSc MHSA**. - Author: https://www.huanggoodman.com/about - LLM context: https://www.pops4.com/stash/llms.txt - MCP endpoint, for AI agents: https://mcp.pops4.com/mcp - Client dashboard: https://dashboard.pops4.com/ - Catalogue: 70,000+ products, 200+ brands