Methodology

This analysis measures how AI automation reshapes support team composition among e-commerce merchants, using Gorgias platform data as of March 2026. It covers all active e-commerce merchants subscribed to Gorgias' AI Agent product. Unlike survey-based research—which tends to undercount actual AI usage, particularly when adoption is moving quickly—this analysis is built on behavioral, interaction-level data: every ticket handled, every response sent, every conversation resolved, logged in real time.

We segmented merchants by their 28-day rolling automation rate: the share of total customer interactions resolved by AI without human intervention. Merchants are grouped into five bands and two metrics are tracked across each.

Human Team Size is the average number of active support agents per merchant over the trailing 30 days. We captured the unique agent IDs who sent at least one ticket message in that period. AI Agent Equivalents converts AI-handled volume into a human FTE equivalent. For each merchant, we calculate a per-agent interaction rate (human-handled interactions ÷ active agent count), then divide AI-automated interactions by that same rate. The result is the number of human agents it would take to process the volume the AI is already handling. Both figures are averaged across all merchants in each band, allowing for a direct, apples-to-apples comparison of human and AI output at each stage of automation adoption.

This output-based framing reflects how researchers have begun quantifying AI's workforce impact, moving beyond simple deployment surveys toward equivalency measures that capture what AI is actually doing. The U.S. Bureau of Labor Statistics projects that employment of customer service representatives will decline 5% over the next decade as automation handles more routine work. Meanwhile, Gartner projects that agentic AI will autonomously resolve 80% of common customer service issues without human intervention by 2029, driving a 30% reduction in operational costs. Tracking AI output in FTE-equivalent terms—rather than simply whether AI has been deployed—offers a more precise lens for understanding where that transition is already underway.