Success story · Backlink mining and SEO · Worldwide
SEOBoostyThree of us built a product that reads up to 1B pages a minute, holds 10B+ links and earns $350k+ a year
SEOBoosty finds backlinks that already exist on sites that never sell links, like The New York Times, Microsoft or MIT, and sells whole link profiles built around them. I was hired as fractional CTO to build the product with a new team of two engineers. I designed the architecture, a global scraping infrastructure of more than 50 scraper servers that reads up to a billion pages a minute, a database that holds more than 10 billion links, and the analytics on top. Crawling at that scale usually belongs to search engines with large infrastructure teams. We run it with three people, a parser that uses no extra memory and infrastructure that scales itself, so reading a million pages costs cents. The first revenue came 3 months after the start, and the product now earns more than $350,000 a year.
- Client
- SEOBoosty
- Product
- Backlinks from sites that do not sell links
- Infrastructure
- 50+ scraper servers around the world, scaling themselves
- My role
- Fractional CTO, hired to build the product
- Team
- 2 engineers and me
- Database
- 10 billion+ links
- First revenue
- 3 months after the start
- Website
- seoboosty.com

In numbers
- pages a minute, read by a product three people run
- 1B
- a year in revenue from a team of three
- $350k+
- links in the database, run by two engineers and me
- 10B+
- to read a million pages: the parser uses 0 MB of extra RAM
- Cents
- scraper servers around the world that scale themselves
- 50+
- from the start of the build to the first revenue
- 3 months
The product
Backlinks from publishers that do not sell links, sold as whole profiles; a link profile no one could build on purpose; a price the publisher sets. Pictures from seoboosty.com.
The starting point
Agencies sell one placement at a time, from sites that sell links. SEOBoosty works the other way. It mines the backlinks that already exist across the web and sells them as whole profiles: one or two links from a publisher that does not sell links, plus up to 1,500 more behind them.
That only works if you can read the web at a scale few companies attempt, from many regions at once, and keep the cost from running away. Search engines do it with whole infrastructure departments. SEOBoosty needed a new product, a new team and infrastructure to match, run by very few people.
What we did
Hired to build it from scratch
I was brought in specifically to build this product with a new team: two engineers and me. I designed the architecture and every system in it so that three people could run what usually takes a data engineering department.
A billion pages a minute from 50+ servers
More than 50 Rust scraper servers spread across regions around the world read up to a billion pages a minute, a volume usually reserved for search engines.
A parser that keeps hardware costs flat
I wrote a fast page parser that processes huge volumes of pages with 0 MB of extra RAM, so more pages never mean bigger servers, and reading a million pages costs cents.
Go, PostgreSQL and ClickHouse
A Go backend over PostgreSQL and ClickHouse holds the pages, more than 10 billion links and the analytics, and stays fast at this volume without a dedicated database team.
Infrastructure that runs itself
An automatic system manages the infrastructure around the world and scales it up and down with the load. The company pays only for the capacity it uses and needs no ops team to watch it.
Analytics that fill the catalog
Analytics over every mined link decide what reaches the catalog: which pages link where, with what anchor, how old the link is and whether it passes authority. Nobody sorts links by hand.
Revenue in 3 months, $350,000+ a year now
The first revenue came 3 months after the start. A customer's whole link profile goes live about an hour after payment, where agencies spend weeks placing links one at a time, and the product now brings in more than $350,000 a year.
The result
Three of us built a product that reads the web at up to a billion pages a minute on more than 50 scraper servers, a scale usually left to search engines. It holds more than 10 billion links, reads a million pages for cents, made its first revenue 3 months after the start and now earns more than $350,000 a year.
Stack
- Go
- Rust
- PostgreSQL
- ClickHouse
Full case · PDF
SEOBoosty
Three of us built a product that reads up to 1B pages a minute, holds 10B+ links and earns $350k+ a year
- 01The architecture that lets three people run web-scale crawling
- 02Scraping up to a billion pages a minute across regions
- 03The parser that does not consume RAM, and what it saves
- 04Automatic infrastructure management and scaling
- 05Backlink analytics: from raw pages to the catalog
- 06The product and how it earns
The full SEOBoosty case
The public story stops here. The PDF has the rest: the architecture before and after, the migration plan, how the team works with AI agents, what it all costs to run and where the savings came from.
More stories
All storiesWant the next story to be about your company?
In a 30-minute call we pick the first task to hand to AI and estimate what it will save you.



