Lesson 76.2 set the bar — a page worth bookmarking. But the most useful guide in the world earns nothing if crawlers cannot find it, phones cannot read it, or no link points to it. This lesson covers the discoverability plumbing: small, checkable hygiene that lets both people and search engines reach your work.

Hygiene, not spells

Say this plainly before any checklist: technical fixes do not rank a useless page. They remove barriers so a useful page can be found, understood, and chosen. A technically perfect empty page still fails the bookmark test. A brilliant page with a broken URL, a "do not index" tag left over from staging, or a title that says "Untitled" fails before anyone judges its brilliance.

Learn the nine terms in non-engineer language:

• Crawl: Google's software visits your page by following links. No link or map pointing to the page means no visit.
• Index: after visiting, Google decides whether to file the page in its library. Blocked, duplicated, or paper-thin pages may be skipped.
• Sitemap: a machine-readable list of your important pages (usually sitemap.xml) that tells crawlers where to look. A courtesy map, not a guarantee.
• Internal link: a link from one of your pages to another. It tells both readers ("go deeper") and crawlers ("this matters, and here is its context").
• Canonical URL: the one official address you declare when duplicates exist (/guide vs /guide?ref=email vs /Guide). It concentrates credit instead of splitting it.
• Robots instruction: a small file (robots.txt) or page tag that permits or blocks crawling and indexing. Essential during staging, dangerous if forgotten at launch.
• Structured data: optional labeled markup (dates, authors, FAQs, events, products) that helps search display richer results. It clarifies; it does not bribe.
• Mobile friendliness: the page reads and taps comfortably on a phone — legible text, no horizontal scroll, buttons far enough apart.
• Performance and accessibility: the page loads quickly, images have descriptions, headings are logical, contrast is readable. Fast and legible helps everyone and everything that evaluates the page.

Diagram the path a page must survive:

published → linked (internal + sitemap) → crawlable (not blocked)
→ indexable (canonical + substantive) → readable (mobile/fast/clear)
→ clickable (title + snippet match intent)

Most "SEO mysteries" are a break in that chain, not a penalty. Check the chain in order before assuming anything exotic.

Search Console: your observation surface

Google Search Console is the free, official window into how Google sees your site. Connect it on day one — it only shows data from after verification, so every week you delay is a week of lost evidence. Three views matter for beginners:

1. URL Inspection: paste any URL and ask "does Google see what I see?" It reports whether the page is indexed, which canonical Google chose, whether crawling is blocked, and what the last crawl found. Use it after every launch or fix. 2. Performance (queries and pages): which queries showed your pages, how often (impressions), how many clicks arrived, and the average position. This is where Lesson 76.1's intent map meets reality — you discover the phrasing real searchers use. 3. Coverage / Pages indexing: which pages are indexed, which are excluded, and why (blocked, duplicate, soft-404, redirected). Treat every unexpected exclusion as a task, not trivia.

Durable idea first, tool layer second: the durable idea is "observe how discovery actually behaves." The Console is today's best free lens for Google; Bing Webmaster Tools is its sibling. Interfaces will shift — the habit of inspecting, reading queries, and clearing coverage errors endures.

Exercise: audit one real page

Pick one public page — your filing guide, a lesson, a local service page — and complete SEARCH-TECHNICAL-CHECK.md:

# SEARCH-TECHNICAL-CHECK.md — [Page URL]
Date: ___  Verified in Search Console: yes/no (property: ___)

1. URL: public, stable, lowercase, no staging tag? Pass/fix: ___
2. Title: under ~60 chars, names the question + audience? Title: "___"
3. Headings: one H1 matching the promise, H2s in logical order? Notes: ___
4. Internal links: ≥1 link IN from your site, ≥1 link OUT to a next step? List: ___
5. Canonical + robots: canonical self-references, no accidental noindex? Result: ___
6. Sitemap: URL present in sitemap.xml? yes/no
7. Mobile: readable on a phone, no sideways scroll, tappable buttons? Screenshot: ___
8. Performance/a11y: loads in <3s, images have alt text, contrast OK? Notes: ___
9. Next step: one visible action (template, booking, signup, lesson)? What: ___
10. Console status: inspected URL → indexed? Queries seen? Coverage issue? ___

How to run it in fifteen minutes: open the URL in an incognito window (catches login walls), read it on your phone, click every link, view the page source and search for noindex and canonical, submit the URL in Console's inspection box, and check the performance tab for any query already surfacing it. Record a fix for every failed line with an owner and date — "fix title by Friday, owner: me" beats "SEO improvements needed."

Finish line: a filled SEARCH-TECHNICAL-CHECK.md plus one verified public URL inspected in Search Console (indexed or with a named fix in progress).

Verify quickly: send the URL to a friend on their phone. If they cannot state the page's promise in five seconds, reach the next step in two taps, or load it before losing patience, the audit found real work regardless of what any score says.

Common failure mode: collecting green scores from a speed plugin while the title still says "Home" and nothing links to the page. Scores are proxies; the chain — linked, crawlable, readable, clickable — is the truth.

Check your understanding

1. What is the difference between crawl and index, and what blocks each? 2. What do canonical URLs and robots instructions each control? 3. Which three Search Console views should a beginner check first, and what question does each answer?

Next

People and crawlers can now reach one good page. Lesson 76.4 asks the scaling question: when should a system generate many pages from data — and when does that factory produce worthless duplicates?