Let's Tilt answers the questions shoppers won't ask a human

Innerwear is a fit category with a privacy problem. The chat data makes the specifics uncomfortably clear. The four recurring themes across 99 days were sagging support adequacy, sizing accuracy from overbust and underbust measurements, no-show claims under fitted clothing, and moisture-wicking and leak absorption performance.

Those are not questions a shopper puts into a live chat window staffed by a stranger, and they are not questions she will wait a day for an email answer to. She either gets an answer in the session or she leaves.

The recorded objections are just as specific: fear that a bra will not lift enough for very sagging or enlarged breasts; uncertainty about the correct size from her own measurements; worry that panty lines or texture will show through clothing; concern about breathability and sweat performance. Every one of them is a purchase blocker, and every one of them is answerable.

The catalog structure compounds it. Revenue lives in assorted multi-packs, while single units sit at a much lower price point. A shopper has to understand pack composition and size consistency across the pack before she will commit to the larger basket.

The Website Agent carries all of it on one surface. AI Sales Chat was trained on the full size grid, the fabric and absorbency specifications, pack composition, and the brand's own guidance on choosing between the seamless, bamboo and leakproof lines.

The privacy of the interaction is the mechanism, not a side effect. A shopper asks the agent what she would not ask a person, gets an accurate answer immediately, and stays in the session. There is no queue and nobody on the other end to be self-conscious about.

The most common opening in the whole dataset was a version of "Help me find my perfect size based on my measurements." The agent takes overbust and underbust figures and returns a size, which is a genuinely useful thing that a size chart on a separate page does not do.

The Nudge Engine covers the rest, shoppers who stall on a product page with a question they never type.

Roughly 87 chats a day on a single channel. The pre-sale ratio is the thing to look at.

Let the agent build the pack. The brand's entry unit and its multi-packs are separated by a large multiple. The agent's job is not to sell a single brief; it is to explain why the assorted three-pack is the right purchase and which sizes to combine. The assorted three-pack seamless brief and the everyday shaper were the two most-discussed products on 99 of 99 days.

Answer fit questions without a queue. Product Info ran to over 5,500 chats and Product Discovery to over 3,000. Order tracking reached only 275 and returns 162 across the entire window. That ratio (enormous pre-sale, minimal post-purchase) is what a well-answered funnel looks like downstream.

Capture size and preference on the first chat. over 4,500 customer profiles were enriched from over 8,000 chats. In a fit-led category, knowing a customer's size and preference before she returns is the difference between a second purchase and a second decision.