# Willy Chavarria Lifted Black Friday Conversion With Autonomous AI Timing Systems *Agentic commerce shifted email sends from batch-and-blast to real-time decision engines that choose product and moment per customer.* By **Jenny Huang Goodman MPA MSc MHSA, Principal** — The Stash Edge, Hako Shikin LLC. Published 2026-10-08. Canonical: https://www.pops4.com/stash/articles/willy-chavarria-2026-10-08t03-1 Subject: Willy Chavarria Tags: agentic commerce, email automation, black friday, personalization, conversion optimization, ai marketing --- Willy Chavarria, the New York menswear label, deployed agentic commerce systems during Black Friday Cyber Monday to automate product recommendations and send timing for each customer, according to Modern Retail. The platform made autonomous decisions about which offer to show and when to send it, replacing the brand's previous batch email schedule with a real-time decision engine that adjusted to individual browsing and purchase signals. The system monitored customer behavior across the site and in prior email interactions, then chose the product, discount tier, and send window without human approval for each transaction. Instead of segmenting customers into buckets and scheduling three or four timed blasts, the brand let the AI determine optimal contact moments based on engagement velocity and cart proximity. Modern Retail reports the approach lifted conversion during the brand's highest-volume weekend, though Willy Chavarria has not disclosed percentage gains. The mechanism works because it solves the timing problem that static email calendars cannot. A customer who browses at 2 a.m. and another who opens at noon need different send windows, and a shopper who added a jacket yesterday needs a different product push than one still exploring categories. Batch sends treat both identically. Agentic systems treat them as separate decision trees. The AI evaluates dozens of signals—time since last visit, items viewed, previous purchase cadence, inbox open pattern—and selects the next best action from a menu the marketer pre-approved. The brand sets guardrails: permissible discount depths, eligible products, maximum contact frequency. The system chooses within those bounds. This is not personalization in the 2018 sense, where a customer's name populates a template. It is decision delegation. The brand teaches the system what a high-intent signal looks like, then permits it to act when it sees one. During a five-day peak like Black Friday, that means thousands of micro-decisions a human team cannot make fast enough. The system sends, waits, measures response, and adjusts the next offer in minutes. A small physical-product brand can run a scaled-down version with existing email tools and one decision rule. Choose your top three SKUs and set a browse-abandonment trigger in Klaviyo or Drip: if a customer views Product A and does not add to cart within **30 minutes**, send a plain-text email with a direct product link and a single sentence about why that item solves a problem they searched for. No discount yet. If they open but do not click within **two hours**, send a second message with a **10 percent** code and a cart link. If they click through but do not purchase within **six hours**, send a final message with free shipping. Each trigger fires based on behavior, not a calendar. You are not choosing send times. The customer's action is choosing them. Set the sequence once, let it run through the weekend, and measure conversion at each step. Cost: zero beyond your existing email platform. Effort: two hours to write three emails and configure the timing rules. The broader pattern is that peak selling windows now reward systems that make more decisions than a human can. A solo founder cannot watch every site visitor and decide who gets which message when. But a decision rule can. The brands that win Black Friday are increasingly those that set the rules, then let the system execute at a speed and scale manual sends cannot match. ## The takeaway Agentic commerce automates send timing and product selection per customer, replacing batch emails with real-time decision trees. --- ## 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