{"id":4999,"date":"2026-03-05T13:00:00","date_gmt":"2026-03-05T13:00:00","guid":{"rendered":"https:\/\/www.rithum.com\/?p=4999"},"modified":"2026-05-14T19:02:52","modified_gmt":"2026-05-14T19:02:52","slug":"etail-west-2026-ecommerce-takeaways","status":"publish","type":"post","link":"https:\/\/www.rithum.com\/blog\/etail-west-2026-ecommerce-takeaways\/","title":{"rendered":"What ecommerce teams are\u00a0focused on\u00a0right now\u00a0in 2026"},"content":{"rendered":"Reading Time: <\/span> 4<\/span> minutes<\/span><\/span>\n

To\u00a0no one\u2019s\u00a0surprise,\u00a0AI talk was everywhere at\u00a0eTail\u00a0West 2026. But the sessions that stuck with me\u00a0focused more on\u00a0returning to basics:\u00a0Product information, checkout, measurement,\u00a0customer\u00a0trust, internal\u00a0readiness\u00a0and meeting the customer where they are\u00a0in\u00a0their\u00a0shopping journey.\u00a0Even as shoppers start their journeys using the\u00a0advanced\u00a0technology of AI,\u00a0those basics\u00a0carry more weight\u00a0than ever. Here’s what I’m still thinking about after soaking up all eTail West had to offer.\u00a0\u00a0<\/p>\n\n\n\n

AI answers are a storefront now  <\/h2>\n\n\n\n

As the\u00a0referral to external websites from LLMs decreases rapidly, retailers and brands\u00a0need\u00a0their products to be the answer when shoppers use AI, not a click along the way.\u00a0Jessyca Frederick, Director of Digital Product at Wine Enthusiast<\/a>, said: \u201cLLM priority #1: answer your questions completely . . . prevent linking out.\u201d\u00a0\u00a0<\/p>\n\n\n\n

Jessyca talked about how these systems gather information: \u201cRAG agents pose fan-out queries to search engines.\u201d RAG stands for retrieval-augmented generation. It means the system searches for sources, pulls what it needs, then writes using what it found. A fan-out query is one question that triggers several searches. Think of one shopper question, many search runs.  <\/p>\n\n\n\n

\u201cThe more fan outs you appear in, the more likely you\u2019re to be cited,\u201d Jessyca said. <\/p>\n\n\n\n

AI answers reuse what they can read and trust.<\/strong>\u00a0Product\u00a0catalog management accuracy\u00a0with an emphasis on data\u00a0hygiene is\u00a0key. The\u00a0fit, price, availability, policies\u00a0must be\u00a0accurate. If your details are incomplete, inconsistent, or stale, the answer engine moves on.\u00a0\u00a0<\/p>\n\n\n\n

Jessyca suggested putting a window for time for \u201cfreshness\u201d\u00a0of that information\u00a0of\u00a03\u00a0months. She emphasized that teams need a\u00a0routine\u00a0for keeping key pages current across the places customers browse.\u00a0The biggest thing to remember is that AI\u00a0bots\u00a0simply\u00a0skip information they\u00a0can\u2019t\u00a0access. Jessyca\u00a0advised to put\u00a0critical product details in formats they can read, keep pages fast, and clean up broken paths.\u00a0<\/p>\n\n\n\n

Shoppers move. Your product story has to beat them there. <\/h2>\n\n\n\n

The \u201ceverywhere customer\u201d panel\u00a0put\u00a0a name to what many teams see in their analytics and support queues.\u00a0With\u00a0Jodi Williams from Ulta Beauty<\/a>,\u00a0Lauren Price from COS<\/a>,\u00a0Feliz Papich from Crocs<\/a>,\u00a0Jason O\u2019Toole from Gildan<\/a>, and\u00a0Jacob Ross from\u00a0PebblePost, <\/a>this panel all\u00a0agreed that\u00a0brands need to stop thinking they can tell consumers where to shop.\u00a0Discovery happens in one place, validation in another,\u00a0purchase\u00a0somewhere else. A TikTok Shop presence <\/a>can end up influencing an Amazon transaction.\u00a0\u00a0<\/p>\n\n\n\n

Instead of focusing on where the conversion came from, the panel advised asking what shaped it<\/em>. Measurement is important, but post-click attribution gives the final click most of the credit, even when the decision was\u00a0made\u00a0three touch-points earlier.\u00a0\u00a0<\/p>\n\n\n\n

\u201cPost click conversions are heavily flawed.\u201d\u00a0according to the panel. Instead, the panel talked about the importance of incrementality.\u00a0Measure lift against a baseline. Set up a holdout, compare outcomes, and see what changed.\u00a0Try\u00a0to reduce guesswork.\u00a0<\/p>\n\n\n\n

Checkout decides revenue  <\/h2>\n\n\n\n

Once discovery is multi-surface and measurement is catching up, it\u2019s tempting to treat checkout as an afterthought. The checkout panel pushed back on that.  <\/p>\n\n\n\n

Pat Suh from Affirm<\/a>\u00a0(joined by Jack Phung from Newegg<\/a> and Henry Spear<\/a> from\u00a0JD Sports<\/a>)\u00a0said,\u00a0\u201cIdeally the AI already knows the\u00a0options,\u00a0but we aren\u2019t there yet.\u201d\u00a0\u00a0<\/p>\n\n\n\n

While agentic commerce grows in popularity to discover products, consumers still abandon carts when checkout creates friction or surprise. Brands and retailers often treat checkout as a problem to solve later, then watch conversion drop after they\u2019ve invested in everything upstream. It\u2019s an expensive way to learn the lesson.  <\/p>\n\n\n\n

Instead, panel members recommended\u00a0that to reduce\u00a0surprises,\u00a0brands should put\u00a0key payment details\u00a0where customers\u00a0look for them,\u00a0make\u00a0choices clear, and\u00a0keep\u00a0terms readable.\u00a0\u00a0<\/p>\n\n\n\n

Personalization needs intent, not a dossier  <\/h2>\n\n\n\n

Anna\u00a0Downs, Digital Personalization Manager at\u00a0The North Face<\/a>, noted\u00a0that\u00a0consumers want to shop without feeling\u00a0monitored. and Michiel Dorjee from Optimizely<\/a>\u00a0said,\u00a0\u201cA lot of people don\u2019t\u00a0like the creep factor of knowing too much about my persona.\u201d The better target, they argued, is intent. \u201cThink about it more of their intent and serve that to them faster, then their experience improves.\u201d\u00a0\u00a0<\/p>\n\n\n\n

Intent is what the customer signals in the moment: what they search, what they click, what they compare, what they\u2019re trying to solve. That\u2019s more useful than demographic profiles and less likely to make someone feel surveilled.  <\/p>\n\n\n\n

Automation raises output, then it raises the stakes  <\/h2>\n\n\n\n

The AI panel with Keri McGhee from Attentive<\/a>,\u00a0George Davis from Cozy Earth<\/a>, and\u00a0Tommy Kowalski from\u00a0HeyDude<\/a>\u00a0covered what teams are truly using AI tools for: scaling creative volume, speeding up lifecycle messaging, and turning one idea into hundreds of variations. The panel cited AI-driven lifecycle journeys producing 200-300% revenue lifts in some implementations.\u00a0\u00a0<\/p>\n\n\n\n

Despite that growth, the\u00a0panel noted\u00a0there\u00a0are still retention risks, because consumers will leave the brand if they have\u00a0a bad experience\u00a0with AI.\u00a0Brands and retailers should understand what automation can do and where a human needs to be involved. Going back to the basics: Customers have always remembered when a problem is handled poorly. AI is no exception.<\/p>\n\n\n\n

Trust keeps working after the purchase <\/h2>\n\n\n\n

There is no replacement for consistency, according to the authenticity and transparency in retail panel.\u00a0Angela Clark<\/a>\u00a0from\u00a0Patagonia<\/a>,\u00a0Catherine Hayden from Kate Farms<\/a>, Elton Graham from Sur La Table<\/a>, and\u00a0Sara Jensen from Hugh & Grace<\/a>\u00a0talked\u00a0about the importance of brand stories that customers can recognize and believe.\u00a0\u00a0<\/p>\n\n\n\n

The big takeaways:<\/p>\n\n\n\n