LLM SEO, shipped and graded
An LLM SEO tool that ships fixes and grades them
Growth Radar is an LLM SEO and GEO tool for sites you can change in code. It checks whether AI crawlers can read each page, ranks fixes such as robots.txt, structured data, llms.txt and answer-first passages by the traffic at stake, opens each as a pull request, and grades the result 28 days after merge.
What it measures, and how
What the tool reads to decide what to fix, and what it does with the answer.
- Readability by AI crawlers
- Whether robots.txt lets GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended in, whether a firewall challenges them, whether llms.txt is valid, and which JSON-LD types each page carries, read as served without running JavaScript.
- A ranked fix list
- One list across every site, ordered by the traffic each fix puts at stake, the lift its kind of fix is expected to bring and the effort it takes. Each item has a playbook: what's wrong, which file to change, the steps, and how to check the result.
- Fixes as pull requests
- For titles and meta descriptions, structured data, page content, internal links, robots.txt, sitemaps, llms.txt and IndexNow keys, Growth Radar opens a pull request on the site's repository, limited to an allowlist of paths per kind of change. It can't merge them.
- A grade on every fix
- The target metric is frozen at merge and measured again 28 days later, against the most similar untouched pages of the same site. Win, loss, flat or inconclusive.
Why an LLM SEO tool should grade its own advice
Most advice about ranking in AI answers is untested on any particular site. A checklist can be followed in full and move nothing. The only way to know is to record what shipped and measure the metric it was meant to move.
Agents in Claude Code or Cursor can work the same fix list through a 41-tool MCP server and record what they ship, so their changes get graded too.
What it can't see
- Shipping needs a GitHub repository; without one, fixes are steps to follow by hand.
- It doesn't write articles or publish to a CMS. Fixes arrive as code changes.
- Anything touching authentication, payments, CI configuration, dependencies or secrets is refused outright.
- On low-traffic pages most verdicts come out inconclusive, because the test can't tell a real change from noise. We show that rather than hide it.
Growth Radar or another tool
Tools that take a different route from finding a gap to changing the site, per their own pages: Checked October 28, 2026.
| Tool | Price | AI engines | Best for |
|---|---|---|---|
| Growth Radar | Free $0 (1 site); Pro $49 a month; Agency $199 a month; 14-day free trial, no card | Narrow: Perplexity, ChatGPT, Gemini and Claude, checked daily | Teams that ship SEO and GEO fixes as code on several sites and want each one graded win, loss or flat 28 days after merge |
| AthenaHQ | Free tier with 300 credits; Starter $295 a month for 3,600 credits; one credit is one AI response | 11, including Grok, DeepSeek, Meta AI and Mistral | Teams that publish new content through a CMS and want the widest engine list at a small-team price |
| Goodie | Core $295 a month (110 prompts); Pro $999 (250 prompts) | 5 on Core, 8 on Pro | Teams that want tracking, crawler analytics, checks of AI claims against their own facts, and writing agents in one bundle |
| Peec AI | Starter $95 a month (50 prompts, 3 models, unlimited seats); Pro $245; Advanced $495; Enterprise custom | Pick 3 models on self-serve plans, each extra one a paid add-on; up to 13 on Enterprise | Marketing teams and agencies that want prompt tracking plus ranked Actions at a self-serve price |
| Scrunch | Starter $250 a month billed annually ($300 monthly), 350 prompts; Growth $417 ($500), 700 prompts | Not stated in the pages checked | Sites on Cloudflare, Akamai or Vercel that want AI bots served a lighter, script-free page at the CDN; Scrunch is now part of Sitecore |
Sources: AthenaHQ pricing · Goodie pricing · Peec AI pricing · Peec AI Actions · Scrunch pricing · Scrunch Agent Experience Platform · Sitecore acquires Scrunch
Common questions
What is the difference between LLM SEO and GEO?
They name the same work: making pages likely to be read and cited by AI assistants built on large language models. GEO, generative engine optimization, is the more common term. The GEO page covers the definition and the instruments that measure it.
What fixes does an LLM SEO tool make?
The ones that decide whether an assistant can read and quote a page. Growth Radar ships titles and meta descriptions, structured data, page content, internal links, robots.txt, sitemaps, llms.txt and IndexNow keys, each as a pull request you review and merge.
Does llms.txt help with LLM SEO?
Whether answer engines use llms.txt when choosing sources isn't established. It is cheap to serve, and agents that read documentation do fetch it, so Growth Radar checks that it's valid and grades a change to it like any other change.
How long does an LLM SEO change take to show up?
Growth Radar waits 28 days after merge before grading, and reads "too early" until then. Crawlers have to refetch the page and answers have to be asked again, so a reading the day after shipping says little.
Can the tool change my site without my approval?
No. It opens pull requests with a fine-grained GitHub token scoped to repositories you choose, and it can't merge them. Every change passes an allowlist of paths for its kind of fix.
Related
Try it on your own sites
Connect Search Console and the last 28 days import straight away. The trial needs no credit card.