What the site had to do

Explain the product. Chatwee is a chat widget you embed on your own site — not obvious from a name or a logo. The homepage states the category plainly and puts scale directly beneath it: registered websites, chat users, messages sent. For a chat tool an empty room is worthless, so proof of use comes before any feature copy.

Show who it's for. Rather than a generic client list, the site names industries — gaming, streaming, radio, trading, online learning, healthcare — so a visitor recognises their own category instead of guessing whether it applies to them.

Make the price legible. Four tiers presented as one comparison rather than four cards: aligned attribute rows underneath, so the reason to move up a plan reads line by line. The free tier is labelled stays that way, forever, which removes the suspicion that kills freemium sign-ups.

Let people try it. The redesign put the live widget on the homepage — real rooms, real users online, a working message box. For a product that sells an experience, that argues better than any feature list.

Remove the fear of installing it. Most visitors who reach the integrations page have already decided; what stops them is the setup. The page answers with the actual script — one line, copy and paste. Developers get their own section below: HTTP API, PHP SDK, JavaScript API, CMS plugins. Two audiences, two paths, deliberately not merged.

‍

DESIGN DECISIONS

• A flat illustration system, held consistently. The client specified the style — flat 2.0, character-led, no gradients or shadows. The work was making it hold across two dozen pages: the same figures, the same limited palette, the same spot-illustration weight everywhere, so the site reads as one product rather than assembled stock.
• Orange reserved for conversion. The palette is blue; every sign-up and trial button is orange, and orange appears nowhere else. On a page that's largely one colour, that makes the action findable without a hierarchy fight.
• Proof placed immediately after the hero. A chat widget nobody uses is worthless, so scale is the argument — and it arrives before any feature copy.
• Pricing built as a comparison, not four cards. Aligned attribute rows beneath the tiers, so the reason to upgrade is legible line by line.
• Two integration paths, separated. A copy-paste script for non-technical users, and API/SDK documentation for developers, kept in distinct sections rather than merged into one instruction.
• A working demo, not a screenshot. The redesign put the live widget on the homepage — the visitor tries the product before signing up, which for a chat tool argues better than any feature list.
• Built to keep growing. The English site I built has since been extended with new pages, new sections and additional language versions — without the underlying system being rebuilt. That's the test a marketing site actually faces.

‍

RESULT

The site is still live years after I left the project. It has been extended since with new pages, new sections and additional language versions, without the underlying system being rebuilt — and the product has grown on it from 50,000 registered websites to over 100,000, and from 70 million messages sent to more than 500 million.

Live URL https://chatwee.com/