SEO
How search engines (and increasingly, answer engines) crawl, rank, and surface content — and what we apply across Pixelbase.
SEO in 2025 is two games at once: ranking in classic Google results AND being cited by AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Claude citations). Both reward the same fundamentals — clean structure, real expertise, substantive content — but the second has a higher bar for specificity and citation-worthiness.
The factors we optimize for
What actually moves the needle across both classic search and AI surfaces.
Indexable rendering
Server-rendered HTML so crawlers see content without JavaScript execution. We use Next.js App Router with SSG/SSR — content is in the initial response.
Structured data (JSON-LD)
Organization, BreadcrumbList, Article, FAQ, HowTo, Product schemas. These power rich results AND help AI overviews understand context.
Internal linking
Topic clusters with a hub page and supporting articles all linking to each other. Distributes authority and helps crawlers understand topical depth.
Core Web Vitals
LCP under 2.5s, INP under 200ms, CLS under 0.1. Real-user metrics from Chrome UX Report drive rankings, not synthetic tests.
Title + meta description
Title 50–60 characters, meta description 150–160. Title gets the keyword early; description sells the click.
Crawl efficiency
XML sitemap, robots.txt, canonical URLs, hreflang for international. Pixelbase auto-generates sitemap + canonical headers per page.
Answer Engine Optimization (AEO)
What makes a page citation-worthy in AI answers
- Concrete answers near the top. The answer engine wants extractable sentences — "The IRS quarterly deadlines are April 15, June 15..." — not throat-clearing.
- Specific numbers and dates.AI models cite pages with verifiable specifics over vague ones. "$300 LLC fee in Delaware" > "low LLC fees in Delaware."
- First-person expertise. Author bios, real names, demonstrable experience. Models penalize anonymous content.
- Q&A formatting. FAQ schemas, H2/H3 headings phrased as questions. Direct match to how queries are phrased.
- Citations of original sources. Linking to IRS, state government, primary research. Signals reliability to both algorithms and AI models.
E-E-A-T (Experience, Expertise, Authoritativeness, Trust)
Experience
First-hand use of what you write about
Expertise
Real credentials and domain knowledge
Authoritativeness
Recognition from peers + citations
Trust
Transparent ownership, sourcing, accuracy
The Pixelbase SEO playbook
- Topic clusters around the decisions every founder makes (entity, EIN, payroll, taxes). Pillar page links to every supporting article and back.
- Server-rendered everywhere. No JS-only content. Crawlers + LLMs see the same HTML you do.
- Schema by default. Organization on every page, Article on each guide, FAQ where appropriate, BreadcrumbList everywhere.
- Quick concrete answers. First 200 words of each guide answer the question. Long-form context comes after.
- OG images per page. Auto-generated with title + subtitle + category. Boosts CTR from social and chat shares.
- Core Web Vitals as a release criterion. Lighthouse score and CrUX data tracked. Regressions block ship.
- Real authors with real bios. Founder + team members named, linked to LinkedIn, with relevant experience highlighted.
What we don't bother with
Keyword density optimization.
Hasn't mattered since ~2013. Write naturally. Use the keyword once in the title, once in H1, naturally in body — that's it.
Doorway pages and thin programmatic SEO.
Google penalizes scaled low-effort content. AI engines won't cite it. Quality + specificity wins.
Buying backlinks.
Paid links from PBNs and link farms are detected and penalized. Build links through actual usefulness — guides others want to cite.
Stuffing content with synonyms or LSI keywords.
Modern semantic models understand topical depth from substance, not keyword variation. Write to teach, not to rank.
Hiding cookie banners and ads behind viewport tricks for Lighthouse.
Real-user metrics are now what matters. Synthetic tests don't reflect what your visitors actually experience.
Quick technical reference
- Sitemap auto-generated at
/sitemap.xml - Robots.txt at
/robots.txt; disallow staging environments - Canonical headers on every page (handles duplicate-content edge cases)
- Image alt text required at upload; OG images auto-generated per page
- Internal link audit runs weekly to catch orphan pages
Related research
Pair this with our Bounce Rate and Intuitive UIs notes — ranking is only useful if visitors stay and convert.
All research notes