KINTO merged two diverging versions of a beverage brand’s site into a single bilingual, English-first codebase, with SEO and FAQs built for US search engines and AI answers.
Problem
A brand entering the US needed a bilingual site with SEO for that market and, later, a direct sales channel: the site showed the catalog but did not allow purchases, and messaging support was manual.
Signals
- Market: A product chosen for its attributes (what it contains and what it doesn’t); target searches in English and Spanish.
- Business: It sells through retailers today; direct sales are the opportunity.
- User experience: Customers had to message, wait and wire money; product pages had to match the real label.
- Support: After-hours questions that “go cold” and sales not measured by channel.
How we understood it
- Full i18n and gap audit.
- Two rounds of client corrections and an SEO audit for the US and for AI answers.
- Commercial proposal built on three frictions: a catalog without checkout, manual replies, unmeasured sales.
What we built and why
- Built: a bilingual site (English first, Spanish as the second version) with a catalog, product pages with real photos and videos, a blog, where to buy and privacy; US SEO, AI-ready FAQs, a sitemap and automatic deployment.
- Proposed, not implemented: an owned store and a messaging assistant with human handoff, on a shared-risk model.
Why this way: Product pages must match the real label and answers must come from verified content; that is why what was built is kept apart from what was proposed, and the store and assistant remain a proposal.
Transformation
-
From two diverging copies of the site to one unified codebase, with both languages corrected and US SEO in place.
-
Images ≈73% to ≈95% lighter.
-
image weight reduction: ≈73–95% depending on stage (measured)
What we don’t know yet
- Whether the store and the assistant get approved.
- No commercial results to report.
