I shut down my SaaS a few weeks ago.
But before that, I accidentally discovered something pretty interesting about SEO.
In ~2 months, Meet Lea went from 0 to 464k SEO impressions.
I wasn’t doing SEO “the proper way”.
No content team.
No expensive backlinks strategy.
No obsession over domain authority.
Most of it came from:
→ programmatic pages
→ comparison pages
→ glossaries / FAQs
→ internal linking
→ link magnets
Basically, I treated SEO more like a system design problem than a marketing task.
What surprised me most wasn’t even the traffic.
It was where the traffic came from.
A lot of users told me they discovered the product through ChatGPT, Claude or Perplexity.
Not through Google directly.
That changed how I think about SEO entirely.
It feels like we’re moving from:
“ranking webpages”
to:
“becoming the source AI assistants trust when answering a question”.
Some pages that barely got clicks on Google still generated signups because LLMs were surfacing them in answers.
That’s also why “buy / compare / alternative to” pages worked absurdly well.
At some point someone literally told me:
“Claude said you were probably the best fit for what I’m trying to do.”
Which is honestly a crazy sentence to receive.
After shutting down the product, I decided to document the entire process while everything was still fresh in my head.
So I turned all my notes into a long playbook.
It covers:
→ the content systems
→ the programmatic SEO setup
→ AI-assisted workflows
→ Claude Code orchestration
→ indexing / crawl issues
→ comparison pages
→ glossary strategy
→ schemas
→ internal linking
→ audits
→ etc
A lot of it is SaaS-oriented, but most of the ideas work for any content-heavy site.
You can give it to your favorite coding agent (Claude / Cursor / Kilocode ...) and it will implement it for you.
Here’s the playbook if you want to read it: