// seo & aeo
Optimize your site for Google AI answers
Check where your website appears in Google AI Overviews and AI Mode, compare cited sources, and get page-specific improvements backed by the evidence. Use the self-serve web checker for citations, or let your AI agent contextualize the results with the CLI. Both use your own Litescrape API key.
You get An AI visibility report, source citations and up to three prioritized page recommendations, exported as Markdown, CSV and JSON.
Find practical improvements for my site in Google AI answers. Use https://litescrape.com/blog/ai-overview-optimizer.md
From one instruction to a useful result
10 questions, two Google surfaces, repeated samples for three questions; up to 40 API attempts in the CLI
- Step 1$0.00
Choose the questions
Optimizer CLI or your agent, 0 calls
Create and review a configuration with your website, brand, category and up to ten buyer questions. Add up to three competitors. The default market is US, English, desktop.
- Step 2$0.003
Observe both AI surfaces
Litescrape Search and AI Mode APIs, 20 calls
Collect an AI Overview observation and an AI Mode observation for each question. Keep mentions, citations, absent overviews and unavailable responses separate.
- Step 3$0.0018
Check answer variation
Same two Litescrape APIs, 12 calls
Repeat the first three questions twice on each surface. Keep repeat samples out of the headline share metrics; they show variation within the run.
- Step 4$0.0012
Compare and prioritize
Public HTML and a bounded retry allowance, 8 calls
Allow at most eight extra attempts within the 40-attempt cap. Read up to ten own pages and ten cited pages, respecting robots rules. Return only recommendations supported by readable evidence, with the facts you still need to confirm.
Example API budget
10 questions, two Google surfaces, repeated samples for three questions; up to 40 API attempts in the CLI
- Choose the questions, 0 calls
- $0.00
- Observe both AI surfaces, 20 calls
- $0.003
- Check answer variation, 12 calls
- $0.0018
- Compare and prioritize, 8 calls
- $0.0012
- Total
- $0.006
Budget at $0.15 per 1,000 successful calls, not a measured receipt. Actual usage depends on matches and available pages. Retries count toward the attempt cap. Agent costs are separate.
What you get back
| File | Contents | Use it to |
|---|---|---|
| report.md | Separate visibility metrics, citations and page recommendations | Choose the next edit |
| observations.csv | Question, surface, sample and mention/citation evidence | Inspect every observation |
| report.json | Configuration, coverage, evidence and reproducible run details | Automate future comparisons |
The web checker and CLI are free software; live runs use your API credits. The web checker reports citations. Your AI agent can use the CLI for deeper interpretation and improvement planning. The offline demo uses clearly labeled sample data.
Put the result to work
Review the recommended edits, verify every proposed business claim, then repeat the same question set after publishing.
A sampled answer is not a ranking guarantee. Crawl failures and unresolved sources stay visible; only compare identical configurations, and do not infer causation from changes between runs.