Find SEO opportunities worth acting on.
Kanso brings together your site's technical analysis, Search Console data and Lighthouse results, turns them into recommendations that show their evidence, and follows each one through action and verification.
Free to start · No credit card required · See how it works
SEO work stalls between the report and the fix
Finding issues is the easy part. Understanding them, deciding what to do and knowing it worked is where the time goes.
Data lives in separate tools
Search performance, crawl reports and speed tests sit in different places, each with its own view of the same site.
Technical issues take investigation
A report says something is wrong. Working out which page, why and what to change is the real job.
Everything looks urgent
Without context about your site, it is hard to tell which findings deserve your limited time.
Generic advice
Recommendations that ignore your pages, your data and your product read like a checklist.
Reports stop at the report
A list of issues is not a plan. It has no owner, no next step and no way to follow the work.
Nobody checks the fix
Once work is marked done, it is rarely confirmed on the live page.
Without Kanso, and with it
- Collecting reports from several tools
- Investigating each issue by hand
- Deciding what matters on instinct
- Turning findings into tasks yourself
- Tracking the work in a separate system
- Checking the live page later, if at all
- Evidence gathered from your site, Search Console, Analytics and Lighthouse
- Findings turned into structured recommendations with reasoning
- You review each one before anything happens
- Approved work becomes a tracked task, or a reviewed pull request for supported fixes
- A re-check of the live page to see whether the fix is there
Evidence, recommendation, action, verification
The SEO Agent is built around a chain, not a report. Each recommendation points at the evidence it came from, each action stays attached to its recommendation, and each fix can be checked against the live page.
Evidence
Facts observed on your site and in your Google data.
Recommendation
A reasoned suggestion that points back to that evidence.
Action
A tracked task, or a pull request you review.
Verification
A fresh look at the live page against a stated condition.
It recommends. You decide.
Kanso does not edit your website by itself, and it does not promise rankings or traffic. It tells you what it saw, why it matters and what could change, and the work happens when you or your team do it, or when you approve a pull request.
From evidence to a verified fix in four steps
Inspect
Kanso collects what it can observe about your site before it recommends anything.
- Analyzes a sample of up to 25 public pages, plus your robots.txt, sitemaps and links
- Connects to Search Console and Google Analytics with read-only access
- Runs Lighthouse through Google's PageSpeed Insights when you ask
- Refreshes Search Console data daily and the public analysis weekly
Prioritize
Evidence becomes recommendations you can read, question and dismiss.
- The AI reads only the facts Kanso assembled, in one explicit run at a time
- Each recommendation names its page, category and reasoning
- Observed data is shown separately from what the data cannot establish
- Wording is labelled as an AI-generated suggestion, never as fact
Act
Approved recommendations turn into work that is tracked.
- An action task moves from open to in progress to marked complete by a person
- Where an action calls for it, the owner approves before work starts
- For supported metadata fixes, preview the exact change and have Kanso open a pull request
- Kanso never merges or deploys. You do
Verify
Finished work is checked against the live page, not assumed.
- Kanso looks at the public page again and tests a stated condition
- Conditions include a title, meta description, canonical link, main heading, structured data, viewport tag, HTTPS and a healthy response
- The result is verified, unchanged or inconclusive
- Tasks without a checkable condition are never called verified
Recommendations you can work through
Active recommendations sit in one list, each with its page and category. Open one to see its evidence, its potential action and what happened next.
What the SEO Agent does
Public site analysis
Samples up to 25 pages for titles, descriptions, canonicals, headings, structured data, image alt text and social tags.
robots.txt and sitemap checks
Reads your robots.txt and sitemaps and flags missing files, blocked pages and sitemaps not declared in robots.txt.
Link analysis
Checks internal and external links for errors, redirects and long chains, and reports the ones it could not check instead of calling them fine.
Search Console insights
Brings in queries, pages, devices and search appearance for the property you choose, refreshed daily.
Analytics context
Adds Google Analytics landing-page and organic data, so recommendations can reflect how pages are actually used.
Lighthouse and Core Web Vitals
Runs Google's PageSpeed Insights on demand and shows Lighthouse results for mobile and desktop.
Evidence-backed recommendations
Each one names its page, its reasoning, the observed data and what that data cannot establish.
Action tasks and pull requests
Track the work as a task with an audit trail, or preview and open a pull request for supported metadata fixes.
Verification
Re-checks the live public page against a stated condition and reports verified, unchanged or inconclusive.
SEO shouldn't stop at a report
Most tools end when the findings are listed. Kanso carries each one forward until it knows whether the fix is in place.
Every recommendation shows its evidence
The facts behind a recommendation are resolved by Kanso from its own data, not written by the model, and each is marked as observed, derived or a limitation.
Changes stay under review
Tasks record who did what. Pull requests change one file with a small, previewed edit, and Kanso does not merge or deploy them.
Fixes are checked, not assumed
A task marked complete is only a person's word. Verification looks at the live page and says plainly what it did and did not observe.
Why automate the legwork of SEO
The judgment stays with you. What can be repeated reliably does not need to be done by hand.
Checks run the same way every time
A crawl and its findings do not depend on who looked or how much time they had.
Evidence stays current
Search Console data is refreshed daily and the public analysis weekly, without a manual export.
Recommendations know your site
They are written against your own pages, your data and your company documents.
Nothing falls through
Each recommendation can become a tracked task, and finished work can be checked.
Kanso SEO Agent, a manual workflow, or a generic AI assistant
| Manual SEO workflow | Generic AI assistant | Kanso SEO Agent | |
|---|---|---|---|
| Project-specific evidence | You gather it from several tools | Only what you paste into the prompt | Your public site, Search Console and Analytics, assembled for you |
| Search Console context | Exported and read by hand | Not connected unless you paste data | Connected with read-only access |
| Technical analysis | Separate crawler and manual review | Cannot inspect your site on its own | Built-in public analysis of up to 25 pages, plus on-demand Lighthouse |
| Structured recommendations | Written up by whoever does the audit | Free-form text | Typed by category, with evidence and stated limits |
| Action workflow | A separate task tracker | None | Tracked tasks, or a reviewed pull request for supported fixes |
| Verification | Re-audit later | None | Re-checks the live page against a stated condition |
| Recurring execution | Whenever someone has time | Only when you ask | Daily data refresh and weekly public analysis |
Tools and workflows vary. This describes the typical case of working by hand, or with a general AI assistant that is not connected to your data.
Part of Kanso, not a separate tool
The SEO Agent works from the same company documents as Kanso's other agents. Start with your website, and see the pricing page for which plans include it.
Got questions?
SEO Agent questions
Where the SEO Agent is used
These use cases name this agent in their workflow. Each one says what it produces and where you stay in control.
Build a Marketing Strategy
Enter your website and Kanso builds the strategy documents, including a marketing strategy and brand voice, that its agents work from.
SaaS
Search, content and community research for software teams whose marketing loses to product work.
Tech Companies
Search and content support for technology companies that sell to developers and buyers.
Ecommerce
Search and content drafts for online stores, built from your own site and Search Console data.
Small Business
Marketing help for owners with no marketing staff: a strategy from your site, then drafts to review.
Law Firms
Practice-area content and search drafts for law firms. Kanso does not check compliance or give legal advice.
Turn your site's evidence into work that gets done.
Kanso finds what is worth fixing, helps you act on it and checks the result. Enter your website to start.
Get started free · No credit card required