English Puzzle
English Puzzle: the teaching method was refined for years. The shop around it was not.
English Puzzle is Anna Kordalewska's own method of teaching English: classes in Warsaw, educational puzzles sold across Poland and through Polanglo bookshops, and daily editorial work (Word of the Day, idioms, study guides). All of it lived in a single WordPress site that had grown for years by bolting on one more thing. We rebuilt it starting from the most expensive part, the path to payment, and then stayed with the site permanently.
- Three businesses inside one site. In-person courses (group, one to one, exam prep, kindergartens), physical product sales and an education blog shared one WordPress install, without shared logic or a shared look.
- The bottleneck sat in the cart and the checkout. The single most expensive moment on the site, the one where the customer pays, performed worst. It was the first thing the client reported and the first thing we took on.
- A theme built up in layers. Product page, category listing and offer pages looked like three different websites, because they had been built at different times with different tools.
- No safe place to make changes. Every fix meant touching the live shop, with no staging environment and no change history in a repository.
- Maintenance pushed to later. Overdue WordPress updates, spam coming through the contact form and slow loading, all treated as something to handle another day.
- Revenue first, aesthetics second. Work was sequenced by impact on sales: cart and checkout before the product page, product page before offer pages, offer pages before the blog.
- No work on production. Before touching anything we set up a development environment and a repository. The client reviews a change before their customers see it.
- One theme instead of five layouts. We rebuilt the templates as a system, so that the shop, the product page, a course page and a blog post speak the same visual language.
- Performance treated as part of selling, not as trivia. In a shop where most traffic arrives from mobile and social, load time is part of the conversion rate.
- We do not disappear after launch. The redesign closed the first stage, after which the engagement moved into ongoing care: updates, incidents, and further work on pages and content.
What we built it with.
Scope of work.
- Foundations
- Development environment and project repository. Local environments for the team and a defined deployment procedure.
- The path to payment
- Cart rebuilt: layout, readability of the order summary, behaviour on mobile. Checkout rebuilt together with its additional elements, released to production as a separate, closed stage. Clean-up of everything the customer sees just before paying, because that is where carts are lost most often.
- Shop and catalogue
- New product page: image, price, promotion and buy button arranged around the decision, not around the description. Rebuilt shop listing and category views.
- Offer pages and templates
- A simple page template and a dedicated page template, two tools instead of five variants. Courses and Pricing pages: five class formats (groups of 4 to 5, one to one, exam prep, conversation classes for mothers with children, kindergartens) presented in one layout with a clear route to contact.
- Content as part of the product
- A Word of the Day module on the home page: a daily reason to come back without opening the blog. Home page blocks and their proportions reorganised.
- Maintenance and performance
- WordPress and plugins updated to a safe version. Contact form spam blocked, on a form that carries course enquiries. Diagnosis and removal of a 500 error. PageSpeed optimisation, first on staging, then on production, continued as part of ongoing care. Day to day fixes after client feedback, handled through a single ticket channel.
What changed.
Deliberately without percentages. We show what actually changed in the client's business and in their shop, rather than numbers we could not honestly attribute to a single change.
Sound familiar?
Your WooCommerce shop works, but it grew for years by bolting things on, and the cart and checkout are the last places anyone looked at. We start where the money is lost, and we stay with the shop after launch.