Works with Claude and ChatGPT
Let AI run your SEO.
Querywright gives Claude or ChatGPT the evidence and the memory to run SEO on your site over time: research, measured search demand, Google’s results, a plan you approve, copy you accept, checks before you publish, and what happened after.
Your AI does the thinking. You decide what ships. Querywright keeps the record straight.
Built for freelancers and small agencies already doing SEO work with Claude or ChatGPT.
This is the App.
Every page, plan, check and result your AI works on is kept here, dated and linked to the evidence behind it.
Overview
What is waiting on you, the state of every page, and this month’s search and traffic, for one site.
Page
Each page carries its own target searches, current work, check result, evidence and results.
Work
The plan in effect, the plans your AI suggested, and the decisions only you can make.
Performance
Rankings, Search Console, Analytics, Local and page experience, each as a dated observation.
Evidence
Every scan, search-demand measurement, Google result and link snapshot your AI has used.
What “run your SEO” means.
Your AI works through the same loop for every site, and comes back to it as the evidence changes. You decide at the points that matter.
- 01
Understand
Your AIScan the site, measure real search demand, and read what Google shows for the searches that matter.
- 02
Decide
YouYour AI suggests which page should own each search. You approve the plan, change it, or say no.
- 03
Create
Your AI, then youYour AI writes from a brief built on the approved plan. You accept the exact copy you are willing to publish.
- 04
Check
QuerywrightQuerywright checks the accepted copy against the plan and the evidence, and names anything that needs your judgment.
- 05
Observe
QuerywrightAfter you publish, Querywright records rankings, Search Console, Analytics, Local and page experience.
- 06
Reconsider
Your AI, then youWhen the evidence changes, your AI brings the decision back to you instead of quietly changing course.
A chat forgets. Your SEO shouldn’t.
What an ordinary chat loses:
- Which research the recommendation came from
- The exact measurements behind it
- What you approved, and when
- The copy you actually accepted
- Whether it was checked
- What changed after you published
What Querywright keeps:
- Every scan, dated, with what was actually observed
- Provider measurements, with their market and date
- Your approvals and acceptances, as you gave them
- The exact accepted copy and its check
- Rankings, Search Console and Analytics, dated
- The full history, for the next conversation
The next conversation starts with the record, not from zero.
Built to be trusted with your site.
AI can make a guess sound like a fact. Querywright keeps observed facts, measured data, your AI’s reasoning and your decisions apart, and says so when the evidence is missing.
- Observed
- Querywright saw this on the site.
- Measured
- A provider measured this.
- Reasoned
- Your AI concluded this.
- Approved
- You chose this direction.
- Checked
- The accepted copy was checked against the plan and the evidence.
- Claim
- “Licensed and insured.”
- Status
- No retained support found.
- What happens next
- Revise the claim, remove it, or leave it for your judgment.
- Zero is a measurement. “No data” is not zero.
- Your AI’s reasoning is never stored as evidence.
- A change after publishing is a sequence, not proof of cause.
Six jobs, one record.
Research and opportunity
Site and page scans, measured search demand, keyword research, Google’s results and competitor pages.
Strategy and content
Plans you approve, writing briefs, the copy you accept, and checks before you publish.
Search performance
Weekly rankings on desktop and mobile, Search Console, and Google’s index reports.
Site quality
What each page actually says and contains, read in a browser when needed, and Core Web Vitals.
Authority and Local
Links, referring domains and link gaps, local pack observations, and your business details by source.
Business outcomes
Organic sessions, engagement, key events you identify as business outcomes, and revenue where Google Analytics reports it.
The demand you measured is the demand the plan cites. The plan is what every check compares against. The rankings are checked for the searches you approved.
Your AI does the thinking. Querywright keeps the work straight.
Work in Claude or ChatGPT as you do today. Querywright connects through MCP and gives your AI the evidence, the plan and the history. The App lets you see and manage all of it.
- Add Querywright to Claude or ChatGPT as a connector.
- Sign in and approve the connection.
- Ask your AI to research your site.
Research https://example.com with Querywright. Identify measured search opportunities and recommend which existing page should own each opportunity before proposing copy changes. Show the evidence and ask for my approval before creating a writing brief.One search, start to finish, on our own site.
Querywright runs on querywright.com. This is the record for one page, as the App keeps it.
- Measured
“google search console mcp”: about 480 searches a month in the United States.
- Observed
Google’s top 50 results for that search, on October 2, 2026.
- Suggested
Your AI proposed a new page to own it.
- Approved
The plan was approved on October 2, 2026.
- Accepted and checked
The copy was accepted, checked, and found on the live page.
- Indexed
Google’s October 5, 2026 report: submitted and indexed.
- Ranking
Not in the top 50 yet, on desktop or mobile.
A new page on a new domain takes time. Querywright records what happened, including what hasn’t happened yet.
$99 a month. Search data included.
- Up to 10 researched sites per billing period
- Up to 50 tracked searches per site, checked weekly on desktop and mobile
- Claude and ChatGPT through MCP
- Cancel anytime
Founding Beta: first ten paying customers, price locked for twelve months. Tax added where required.
See pricing