HubSpot SEO II Certification cheat sheet

Silo architecture: each topic cluster gets a pillar page plus supporting pages that link up to the pillar and laterally to close siblings; cross-silo links dilute topical focus.

Crawl depth rule: any page a user needs should be reachable within 3 clicks of the homepage; pages deeper than 4 clicks are crawled less often and rank with less stability.

Link equity is split, not duplicated: a page with N internal outbound links passes roughly its own equity/N per link, so navigation bloat (mega-menus, tag clouds) lowers per-link value.

Orphan pages (zero internal inbound links) are indexed only via sitemaps or backlinks and are the most common cause of 'indexed but never ranked' pages.

URL structure decision rule: prefer flat, hyphenated, lowercase, keyword-bearing slugs; nesting beyond 3 path segments adds no ranking value and complicates migrations.

Redirect rules: 301 for permanent moves and consolidations; 302 only for temporary tests; never chain more than one hop and never redirect to a 404 or to a noindexed URL.

Canonicalization: one canonical URL per piece of content; self-referencing canonicals on all indexable pages; parameterized or duplicate variants canonicalize to the clean version.

Pillar/cluster ratio heuristic: 1 pillar to 8-15 supporting pages keeps the pillar's internal link count manageable and the cluster's topical signal coherent.

Site migration checklist order: crawl baseline, map old-to-new URLs 1:1, deploy redirects, keep sitemaps current, monitor 404s and rankings for 4-8 weeks.

Search intent taxonomy: informational, navigational, commercial investigation, transactional; the SERP's dominant result type defines the intent you must match to rank.

Keyword difficulty is a proxy, not a threshold: compare the referring-domain strength of the current top 10 against your own domain before committing to a head term.

Long-tail math: a term with 200 monthly searches and difficulty 5 usually outperforms a 20,000-search term at difficulty 70 for a young domain in total traffic per unit of effort.

Topic clusters beat isolated keywords: one page per intent, not one page per synonym; near-duplicate keywords cannibalize each other and split internal links.

Keyword cannibalization test: if two URLs alternate ranking for the same query in Search Console, merge them or differentiate their intent explicitly.

Content gap analysis: subtract competitor ranking keywords from your own ranking set; the remainder sorted by volume × intent value is the editorial roadmap.

Search volume is a 12-month average in most tools; apply a seasonality multiplier before forecasting traffic for seasonal terms.

Zero-volume trap: low-volume long-tail queries often appear as zero in tools but carry real demand; validate with Search Console impressions or autocomplete.

Title tag: 50-60 characters to avoid truncation; primary keyword early; each page's title must be unique across the site.

Meta description: roughly 120-155 characters; not a ranking factor but drives CTR, which affects click-through signal quality; write it as ad copy with the query's intent.

Heading hierarchy: exactly one H1 matching the page's primary intent; H2/H3 for subtopics; headings are a structural signal, not a keyword-stuffing slot.

Image SEO: descriptive filenames, alt text describing the image's content, explicit width/height to prevent layout shift, and modern formats (WebP/AVIF).

Internal anchor text should describe the destination, not say 'click here'; varied but descriptive anchors distribute relevance and avoid over-optimization patterns.

E-E-A-T signals on-page: named author with credentials, cited primary sources, last-updated date, and contact/about pages linked from the article.

Featured-snippet targeting: answer the query in a 40-60 word paragraph directly under a question-style H2, then expand with detail below it.

Content freshness rule: update pages whose rankings decay and whose facts are stale; do not change the URL, and log the update date visibly.

robots.txt controls crawling, not indexing; a Disallow keeps a URL out of the crawl but a linked URL can still be indexed without content. Use noindex to keep pages out of the index.

noindex plus Disallow is a conflict: if the page cannot be crawled, the noindex is never seen and the URL may still be indexed.

XML sitemap limits: 50,000 URLs or 50 MB uncompressed per file; use a sitemap index for larger sites; list only canonical, indexable, 200-status URLs.

Core Web Vitals thresholds (75th percentile, field data): LCP ≤ 2.5 s, INP ≤ 200 ms, CLS ≤ 0.1.

Crawl budget is a function of site authority and server response time; a slow TTFB reduces crawl rate more reliably than any other single factor.

Status code semantics: 200 indexable, 301 permanent, 302 temporary, 404 gone, 410 gone permanently and crawled faster than 404, 5xx server error (retried, not deindexed).

JavaScript rendering: content injected only by client-side JS may be missed or delayed; prefer server-side rendering or dynamic rendering for indexable content.

Hreflang requires reciprocal return tags; a missing return link invalidates the pair and can cause the wrong locale to rank.

Duplicate content resolution order: canonical tag, then 301, then parameter handling in Search Console; never rely on the canonical alone for large-scale duplication.

Search Console metrics: impressions, clicks, CTR = clicks/impressions, and average position; average position is an average of positions, not the rank of your best keyword.

Position 1 typically earns roughly 25-35% CTR, position 2 roughly 12-18%, position 3 roughly 8-12%; use these bands to estimate click upside from rank gains.

Organic traffic attribution: direct traffic often absorbs organic sessions when tracking parameters are stripped or when users return without a referrer; expect 10-20% undercounting.

Segmentation rule: always split branded vs non-branded queries; branded growth inflates totals and hides whether SEO actually acquired new demand.

Rank tracking is a sample, not a census: tools use fixed locations and devices, so treat movements of 1-3 positions as noise unless sustained across a week.

Conversion attribution: last-click undervalues upper-funnel content; use a first-click or linear model to judge informational pages' contribution.

Reporting cadence: weekly for technical alerts (5xx, crawl errors), monthly for rankings and traffic, quarterly for content ROI and cluster-level performance.

Baseline before you change anything: record impressions, clicks, CTR, and average position per URL so post-change deltas are measurable.