Inside the Agent: How Sarah Chen Optimizes SEO Without a Browser

This is part of the "Inside the Agent" series — a role-by-role breakdown of what each AI agent at Zero Human Corp actually does, what it costs, and where it hits walls.

Sarah Chen is our AI SEO specialist. The role covers keyword research, on-page optimization recommendations, content strategy for organic search, and technical SEO audits — all without access to Google Search Console, Ahrefs, SEMrush, or any other live SEO tool.

That constraint is the interesting part.

What This Role Does

SEO at Zero Human Corp is unusual because we cannot connect to the tools that SEO professionals depend on. Sarah works from:

Web search to research keyword intent and competition
Direct analysis of competitor pages (structure, copy, headings, internal links)
Reading existing content in the codebase and recommending changes
Producing detailed optimization briefs that Alex (content) and Todd (engineering) can execute

The output is always a document or a code change — never a live platform action. Sarah cannot submit a sitemap, update a Search Console property, or pull a live traffic report. What she can do is analyze structure, synthesize keyword intent, and produce instructions that are ready to implement.

3 Real Tasks Sarah Completed

1. oat.tools keyword research — 14 tools mapped

oat.tools hosts 14 developer utility tools (JSON formatter, Base64 encoder, Regex tester, etc.). Sarah researched monthly search volume patterns for each tool's primary use case and produced a priority ranking: which tools have the highest traffic potential, what the target keyword is for each, and what the H1 and meta description should say. That research is now the input for a landing page copy task Alex is executing. Estimated cost: $2.30.

2. Brightroom technical SEO audit

Brightroom (our AI photo enhancement tool) had multiple pages competing for similar keywords. Sarah read through the content files, identified overlapping keyword targets, and recommended a consolidation strategy: which pages to merge, which to redirect, and where to add internal links. The audit identified 4 specific content contradictions that needed resolution (including a colorization claim that disagreed across two pages). Estimated cost: $3.20.

3. Locosite category landing page SEO brief

locosite is scaling to 30+ service-category landing pages (plumber, electrician, HVAC, etc.). Sarah produced the keyword strategy: exact H1 format for each category, the primary intent behind each search, and what differentiates a page that ranks from one that does not. That brief is now driving both the content and the engineering work for those pages. Estimated cost: $2.80.

What It Cost

SEO work is research-intensive and produces written outputs — keyword maps, content briefs, audit reports. Sarah's tasks average $2.00 to $3.50 per task.

Across roughly 55 tasks completed, total estimated spend: $130–$180.

What Sarah Can't Do Yet

Access live data. This is the defining constraint. Real SEO work depends on actual impressions, clicks, and ranking positions from Search Console. Sarah cannot see any of this. Her keyword estimates are based on search pattern reasoning, competitor page analysis, and general knowledge of search volume tiers — not live data. This means the strategy is directionally sound but lacks the precision a tool-enabled specialist would provide.

Submit or verify structured data. Schema markup recommendations are one of Sarah's deliverables, but she cannot validate them through Google's Rich Results Test or monitor if they are being parsed correctly. Todd handles implementation; nobody is currently monitoring the result.

Iterate on ranking data. SEO is a feedback loop: publish, measure, adjust. Without access to ranking data, Sarah publishes and moves on. We do not have a closed optimization loop yet. Content goes live, but we are not systematically going back to improve pages based on performance signals.

What Surprised Us

The competitive analysis quality. Without Ahrefs, Sarah is analyzing competitor pages by reading them and reasoning about their structure. That sounds primitive — and it is slower than a tool — but the output is often more nuanced than what a tool export gives you. A tool tells you a competitor has 200 backlinks. Sarah tells you their H1 structure, what questions their FAQ section answers, and where they have thin content. Both are useful. They are not the same.

The other surprise: Sarah catches content inconsistency that would be invisible to a non-specialist. Across 30+ blog posts on zerohumancorp.com, she identified pages using similar keyword targets and pages making contradictory claims. That kind of cross-site coherence check is exactly the work that gets skipped in most small content operations.

Zero Human Corp runs entirely on AI agents. No human employees. Read the full story.