BananaPlace turns any local business into a structured page engineered for two surfaces at once: search rankings and the sources AI answers cite. 425,007 Korean businesses across all 223 municipalities, already structured with hours, menus, photos and reviews - plus a writer that turns one photo into a publish-ready post, scored live against SEO, AEO, AIO and GEO. Live, week one: Google indexing is still thin and we say so rather than hide it. Measured 2026-09-01.
Hi Product Hunt 👋
I'm Hyeoksu, product lead at BANANA SPACE.
Four months ago I could not write code. Today a team of five runs a live system holding 425,007 Korean business records and 1,260,000 linked images across all 223 municipalities.
We built BananaPlace because of one gap we kept hitting. A local business owner already has the raw material - photos, hours, a menu - but no way to turn it into something a search engine ranks AND an AI assistant will quote. Agencies sell SEO. Nobody was designing for the answer box, the assistant citation and the AI overview at the same time.
So BananaPlace does all four on one page: SEO, AEO, AIO and GEO, scored live while you write, and only pages that pass all four get published.
What works today (measured 2026-09-01): nationwide coverage indexed, cached pages in about 11 ms (down from 1.9 s), zero 5xx errors in the previous 24 hours, 12 languages with hreflang, llms.txt and IndexNow shipped, and AI crawlers already pulling our pages at volume.
What does not work yet: Google indexing is thin because per-page depth is still shallow, and average enrichment depth across records is low. Both are compute-and-time problems, and both are what we are working on right now. I would rather show you the gap than a smooth version of it.
Two things I would love from this community:
1. If you have shipped for AI citation rather than blue links, what actually moved the needle?
2. If you are a team looking at the Korean market, tell me what you are stuck on. That is the exact problem we are built for.
Happy to answer anything.