Keep search and content moving while your team ships the product.
Kanso reads your Search Console data and your public site, drafts articles for the queries where you are not yet on page one, and checks how readable your product pages are to AI systems. You review and publish everything yourself.
Free to start · No credit card required. See how it works
At a glance
Marketing is what slips when the roadmap is full
A software company changes its product constantly, and its public pages must keep up. The people who could do that are the people building the product.
Product pages drift out of date
Feature, pricing and integration pages each need a title, a description and a canonical link, and they change with every release.
Your own search data goes unread
Search Console shows queries where you sit just off page one. Turning one into an article is a project, so it waits.
Docs and pricing must be readable to machines
A redesign can leave a robots.txt rule or missing llms.txt that blocks AI crawlers, and the pages still look fine.
From scattered checks to a short list you can act on
Kanso does not run your marketing. It does the reading and the first draft, so your time goes on decisions.
- Hunting through Search Console for a query worth an article
- Finding out about a blocked crawler or a missing tag by accident
- Searching Hacker News by hand for your category
- One qualifying query chosen from your own data, and a draft written for it
- A readable status for crawler rules, llms.txt, structured data and your core pages
- Recent Hacker News conversations collected with the reasons they matter
Three agents carry the work, one adds research
The first three cover your site and your content. Hacker News looks outward, at where your category is discussed.
SEO Agent
Analyses up to 25 public pages, such as pricing and feature pages, and turns Search Console data into recommendations with evidence.
- It produces
- Findings and tasks for the pages that carry your product.
- Where it fits
- Reads the site and the search data first.
Writer Agent
Picks a Search Console query where you are not on page one and writes one long-form article draft in your brand voice.
- It produces
- A Markdown draft in your Agents Feed.
- Where it fits
- Turns search demand into content.
GEO Agent
Checks nine named AI crawlers against your robots.txt, looks for llms.txt and structured data, and estimates how citable your pages look.
- It produces
- A readiness analysis and a labelled estimate.
- Where it fits
- Checks how machines read the site.
Hacker News Agent
Searches the keywords you set, such as your category, for conversations worth your time.
- It produces
- Research notes, never finished text.
- Where it fits
- Shows where your category is discussed.
A working loop for a product team
Six steps. Kanso runs the reading and the drafting; you run the decisions.
- 01You
Start from your site and your search data
Enter your website and Kanso writes five company documents from it. The workspace owner then connects Search Console, read-only, and chooses the property.
Read the pages that sell the product
Kanso samples up to 25 public pages and checks titles, descriptions, canonical links, headings, links, robots.txt and sitemaps. Findings become recommendations with the evidence behind them.
Turn one search opportunity into a draft
From the last 30 days of queries, it keeps those with more than 50 impressions where you are not on page one, picks the most searched, and writes one article. With no qualifying query, it writes nothing.
Check what AI crawlers can read
It reports whether each named crawler is disallowed, whether llms.txt exists, and how your structured data looks. A model also estimates citation potential and labels it as an estimate.
Look at the conversations
Run the Hacker News Agent for stories and comments from the last 48 hours that match keywords you set, such as your category.
- 06You
Review, publish and decide
Read the notes and the draft, edit what needs your product knowledge, publish it in your own tools, and make any site change yourself or with your developers.
What you receive
Everything waits in your Kanso workspace until you act.
An article draft
Long-form Markdown aimed at 1,400 to 2,000 words, with its target query shown. The length is requested, so it can come out shorter or longer.
Site recommendations
Each names its page, its reasoning and the data behind it, and can become a tracked task.
An AI-readability check
Pass, warning, missing or not measured for each signal, plus an estimate labelled as one.
You publish and you change the site
Nothing in this loop touches your website, your blog or any community.
Kanso does
- Reads your public site and your search data
- Drafts one article per run and writes research notes
- Records the tasks you open
You decide
- Whether a query is worth an article, and what the final article says
- Where and when to publish
- Which recommendations become work, and who does it
Kanso never
- Publishes an article or edits your site
- Posts to Hacker News
- Posts a Reddit reply you have not chosen to send
Where this fits
Situations, not results. What Kanso finds depends on your own site and queries.
After a site redesign
Run the analysis to see whether titles, canonical links and crawler rules survived, and whether llms.txt is still served.
A feature people search for by problem
If Search Console shows that problem on page two, the Writer Agent can draft an article for you to rework.
A category thread on Hacker News
Research notes show what a useful comment could address, for you to write in your own words.
How Kanso approaches software marketing
Each point is something the product does today.
It starts from your own data
The article topic is a query your site already appears for, and the findings are about your actual pages.
Evidence travels with the advice
Recommendations show what was observed and what the data cannot establish, so a developer can check them.
One context for every agent
Your product information and brand voice are what the writing agents draw on.
What Kanso does not do for SaaS teams
A software marketing stack is wider than these agents.
It does not publish to your blog or docs
There is no CMS integration. You copy or download the draft and publish it yourself.
It does not write on request
The Writer Agent chooses from your Search Console queries. You cannot ask it for a comparison page or a release post.
It does not see your product
No in-app events, sign-ups, churn or pipeline. Kanso reads your public pages and search data only.
It does not track rankings or AI answers
It keeps no daily keyword history and never asks an AI product about you.
Related use cases
Turn Search Console Data Into Content
Turn Search Console opportunities into long-form article drafts in your brand voice, for you to review and publish.
The Writer workflow in more detail.
Get Found in AI Search
Check how readable your site is to AI systems, estimate how citable it looks, and turn the gaps into actions.
The GEO checks in more detail, including what they cannot measure.
Find Reddit & Hacker News Conversations
Find relevant Reddit threads and Hacker News conversations. Reddit replies are drafted for you to post; Kanso never posts to Hacker News.
How Hacker News and Reddit research works, including how relevance is judged.
Startups
Research launch conversations and draft posts when there is no marketing hire yet.
For teams earlier than this, with no marketing hire yet.
Questions about this use case
What does Kanso do for a software company?
It reads your public site and Search Console data, drafts articles from queries where you are not on page one, checks how AI crawlers read your pages, and researches Hacker News. You review and publish.
Will the Writer Agent work without Search Console?
No. Without a connected Search Console the Writer Agent stops. The site analysis and the AI-readability checks read your public pages.
Can Kanso write a comparison or integration page?
Not on request. The Writer Agent picks its own topic from your Search Console data and takes no brief.
Can it look at my product or my customers?
No. Kanso reads your public pages and your search data, not your product, your users or your sales pipeline.
Does it publish to my blog?
No. It has no CMS integration. Drafts are Markdown you copy or download.
Does Kanso tell me whether AI assistants recommend my product?
No. It checks the signals your site exposes and estimates citation potential. It does not ask any AI product about you.
See plans and pricing · All use cases
Give your site the attention the roadmap leaves no time for.
Enter your website. Kanso reads it, finds what is worth fixing or writing, and leaves every decision with you.
Get started free · No credit card required