SEO for an online school: getting students from search without ads

Paid traffic disappears the second you pause the campaign. Search works differently: an article you published in March still brings leads in December. Here is a working plan to build an SEO channel for an online school in 3–6 months — no agency, no link budget.

What it actually returns

The key difference from ads: cost per lead goes down over time instead of up. If an article took you 4 hours and brings 150 visits a month for two years, that is 3,600 visits from a one-off investment.

The downside: for the first 2–3 months you will see nothing. Search engines evaluate new content slowly, and a young domain still has to earn trust. SEO is not a replacement for ads — it is the second floor. Ads bring sales now, search brings a baseline flow in six months.

Planning rule: 1,000 visits to an article yield 20–40 email or bot subscribers (2–4%), and 1–3 sales from those if your funnel works. It only compounds at volume — 20–40 articles, not three.

Three types of queries

Problem-aware (TOFU). The reader does not know your product exists yet: "how to record video at home without a studio". High traffic, 0.5–1% conversion to sale. The job of this article is not to sell but to capture a contact: a checklist, a template, access to a bot.

Solution-aware (MOFU). The reader is looking for a way: "online nutrition course", "platform for online courses", "how to accept payments for a course". Conversion 2–5%. The most valuable type — and the one experts usually skip.

Brand and comparison (BOFU). "CREO vs Kwiga", "Zenedu alternative", "reviews of course X". Least traffic, but these people have a card in hand: conversion up to 10%.

The classic mistake is writing only TOFU "useful content" and having zero BOFU pages. People read your blog and buy from the competitor who built that page.

Keyword research in one evening

No paid tools needed at the start:

1. Type your main topic into Google and write down every autocomplete suggestion plus the "People also search for" block — that is 30–50 real queries.
2. Open 5 competitors and list the headlines of their ranking articles.
3. Go through your own conversations: consultation questions, Telegram comments, objections in DMs. That is the language of your audience, and no tool has it.
4. Put it in a table: query — type — rough volume — priority.

Do not chase head terms like "online courses". A young site will not show up there. Take long queries of 3–6 words: 50–200 searches a month, but you can realistically rank, and the intent is far sharper.

Article structure

• One H1 with the keyword per page.
• The first paragraph answers the query in 2–3 sentences. If someone searched "how much does it cost to launch a school", the number belongs above the fold.
• H2 subheads for adjacent queries — one article then ranks for 10–15 keywords instead of one.
• 900–1,500 words for guides, 500–700 for narrow answers. Longer is not better.
• Specifics: numbers, checklists, real examples. A paragraph with no fact in it is a deletion candidate.
• A 3–5 question FAQ at the end — a shot at the answer box and a way to close remaining objections.
• 3–5 internal links to your own material.
• One CTA, not five.

Technical basics

• unique title (under 60 characters) and description (under 155) on every page;
• sitemap.xml and Google Search Console connected — otherwise you are flying blind;
• speed: LCP under 2.5 seconds, images in WebP;
• mobile version — 70–80% of blog traffic arrives on a phone;
• Article and FAQPage structured data;
• on a bilingual site, correct hreflang and canonical URLs so versions do not compete.

What to measure

Check four things in Search Console weekly: impressions, CTR, average position, indexed pages. Impressions growing but no clicks means the title and description need a rewrite. Position 11–20 means the article is stuck on page two: add a section, refresh the data, add two internal links, and it often moves within a fortnight.

On the school side, measure what matters: how many organic visits became subscribers, and how many became payments. Traffic without those two numbers is a vanity metric.

A realistic plan

24 articles in 3 months (two a week) is the volume at which the channel starts paying back. Split: 60% MOFU, 25% TOFU, 15% BOFU. First results in weeks 8–12, plateau at 6–9 months. After that, updating old material outperforms publishing new — at half the cost.

Three mistakes that cost a year

1. Writing about yourself instead of the query. Nobody searches for "how I got into this profession".
2. Quitting after five articles. Five articles is not a channel, it is a test.
3. Nowhere to send the traffic. Visitors arrive and there is no form, no bot, no sales page.

That is why a search channel only makes sense alongside the rest of the infrastructure: a sales page, a subscription, a funnel, and a platform that carries students to the end of the course. CREO puts those in one place — a sales page builder, Telegram funnels, CRM, payments via WayForPay and Stripe, and gamification that keeps course completion at 70–80%. Beta testing starts in August 2026 — join the early list at platform.creo.ua.