The problem was not a lack of programs. It was that a parent could not tell which format fit: individual or mini-group, kids or adults, exam or conversation. When the answer lives in a DM, the parent leaves.
We started with the questions, not the design: who is the student, what level, how often, what is included, can they try it. Each program got its own URL and its own page. Prices moved to a dedicated page. The trial lesson became the primary CTA.
The blog was built around parent questions, not generic “10 grammar mistakes” content. Every article links to a program and the trial lesson, so content supports the commercial pages instead of competing with them.
The English section of that site later became a testbed for the same approach we are using here: language-specific content, clear pricing and a lead form that asks for the minimum.