SEO research system
Rankr
An AI SEO agent for Singapore SMEs that finds demand and competitor gaps, drafts from real audience questions and helps businesses show up on Google.
Our checker gave Lullaflex scores between 93 and 100. Google still ignored it. That failure changed what Rankr measures.
- Product demo: Rankr researches competitor gaps, drafts the useful page and waits for approval before publishing.
- ChatGPT citation: The low-authority Lullaflex testbed earned a real citation from ChatGPT.
- UCWS pitch: Rankr reached the Top 20 from more than 800 UCWS participants.
- UCWS finalist: Rankr reached the Top 20 from more than 800 UCWS participants.
- SME benchmark: The benchmark compares 42 responding sites by sector instead of hiding them behind one score.
- Live testbed tool: The live testbed included a real planner because useful tools earn links and citations that commodity articles do not.
What it does
Rankr combines market research, site diagnostics and content planning. Each recommendation links back to the evidence that produced it, and the public checker gives a small business a useful result without a sales call.
The result
Two Singapore SMEs signed letters of intent before we wrote any code, and their discovery calls shaped the first version. Rankr later reached the Top 20 from more than 800 UCWS participants. Each finalist team received US$10,000 in OpenAI credits, and the project received S$2,000 from SUTD's Baby Shark Fund.
The experiment that changed the product
We tested Rankr on Lullaflex, a brand-new mattress site. Its pages scored between 93 and 100 on our checker, but Google still barely showed them because the domain had almost no authority. Our score was telling us the pages were tidy, not that anyone could find them.
That changed the product. Rankr now separates page fixes from search demand and authority instead of hiding everything behind one number. The same DR 2 test site was later cited by ChatGPT as "Best for Singapore", which was stronger evidence than another score we invented ourselves.
How it works
The agent starts with competitor gaps and real Reddit questions instead of a blank prompt. It remembers what each client has already published so it does not pitch the same topic twice.
I built the client sites, publishing layer and the knowledge graph behind that memory. Long research jobs run on Trigger.dev, and nothing is published until a person approves it.