All Posts

"Can it recommend the right product automatically?" Founders ask me this a lot. It's the right instinct and slightly the wrong question. Automatic is easy. Every tool on the market can fire a recommendation the moment someone opens a chat. Bestsellers. Frequently bought together. What people like you bought. That's not a recommendation, it's a guess dressed up in a carousel. A customer walks into an electrical store and says, "I need an inverter." The guy behind the counter d

Every AI assistant demos well. That is the problem. Book five demos this week and you will watch five near-identical performances: a customer asks a question, the agent answers beautifully, everyone nods. Nothing you saw in those thirty minutes tells you which one will still be earning its keep in six months. The demo is a scripted conversation on a clean catalogue with a friendly question. Your customers are none of those things. So the decision has to be made on things the

A founder asked me this last week. I went looking properly. Here's the honest map. The support-first camp. Gorgias, Zendesk, Intercom's Fin, Zowie. Built as helpdesks, with AI bolted on later. Gorgias' AI Agent now plays two roles — support for returns, tracking and cancellations, and a shopping assistant before purchase. Strong at deflecting tickets. Selling is the newer muscle. The commerce-first camp. Rep AI, Alhena, Envive. Built to convert browsers on the storefront, wit

Three questions kill more D2C sales than price ever will. Not because they're hard. Because nobody answers them in time. 1. "Will this actually work for me?" An efficacy question. The customer isn't doubting your product — they're doubting whether their situation is the exception. Sensitive skin. Hard water. A body that reacted badly to the last three things they tried. A product page can't answer this. It doesn't know their situation. So it says "works for all skin types" an

Or: what I learned about D2C conversations from watching an AI agent do the job my sales team was asleep for. Setup: A skincare brand we work with does about 40% of their WhatsApp inquiries between 9 PM and 1 AM. Their support team logs off at 8 PM. For years, this was the deal. Late-night messages sat unread until morning. By the time anyone replied, half the customers had moved on, bought elsewhere, or forgotten why they messaged in the first place. They knew they were losi

For the last five years, "conversational commerce" mostly meant one thing: a chatbot on your website that no one used, and a WhatsApp broadcast tool that annoyed your customers into unsubscribing. That era is ending. Fast. Here's where I think the next 12 months take us: 1. Bots die. Agents take over. The chatbot layer that D2C brands bolted on in 2022 was decision-tree software wearing an AI costume. Customers figured it out in one message. The next wave is different — agent