01 · Case study · 2026
tonweb.pro: Conversational & AI UX
Explore a website subscription by reading the offer or asking its AI guide
Live websiteVisit tonweb.pro
- Role
- Founder, Product Designer & Front-end Developer
- Timeline
- 2026
- Industry
- WaaS · Local Business · AI-guided UX
An AI guide and in-page explainers let visitors explore the offer through open conversation or precise questions right where they arise. The conversational experience is built around tonweb.pro’s website subscription for local craftspeople and small businesses in Wallonia: three plans, no setup fee, and clear explanations of what is included, how unused hours work, and what happens after 24 months.
A website subscription is hard to grasp in one look. Craftspeople have little time and even less patience for jargon, and the usual offers hide what matters: a setup fee in disguise, a contract that is unclear at the end, hours of changes that vanish if unused. Market research, competitor analysis and conversations with artisans pointed to the same finding: people do not leave because the price is too high, they leave because they cannot picture what they would actually get, and nobody is there to answer the question they are too shy to ask. The brief became a comprehension problem before a design one: make the concept obvious, the pricing almost boringly simple, and never leave a visitor alone with a doubt.

The work started with user research and competitive analysis, then turned every recurring doubt into a design element. Pricing is three plans on a single selector, one price each, with the 24-month total shown and no setup fee, so nothing has to be calculated. The least intuitive rule, unused modification time carried into a reserve capped at 12 months, is not written as a paragraph: it is a scroll-driven gauge that fills month by month and stops at its cap, per plan. On top sits an AI layer that is more than a chatbot. It has two parts: a classic assistant for open conversation, and small in-page explainers placed exactly where visitors usually stall, on each plan line, each guardrail, each FAQ answer and each local search intent. One tap asks a precise question, the answer opens right beside it, and the exchange is kept and can continue in the chat. The AI is trained on the company’s commercial strategy: it knows the products, the pricing logic and the rules, recommends a plan from a description of the visitor’s activity, explains, and hands over to the team when a human is needed. Its knowledge base is rebuilt from the same data files as the pages, so editing a price updates the site and the assistant together. Chat streaming never auto-scrolls, so people read at their own pace.
The result is a site that can be navigated in two ways: by reading, or by asking. The concept is understood without a sales call, the pricing reads in seconds, and the AI works as a guide that lifts the questions people usually take away unanswered. The offer is also built to be found: local pages for 12 towns and 6 trades, structured data and machine-readable files for search engines and AI assistants.
Drag or swipe to browse, or focus the screen browser and use the arrow keys. Press Enter or select Full size to enlarge the current screen.
Drag to explore · Click to enlargeSwipe to explore · Tap to enlarge




