A price platform built for scale
The public platform behind one of the web’s largest menu-price datasets — turning 70M+ menu items across 92K+ brands and 709K+ locations into fast, searchable pages that stay quick under heavy traffic and rank across millions of URLs.
-
I Brief A goldmine of data, no fast product around it
PriceListo sits on one of the largest collections of menu and service prices anywhere — tens of millions of items across restaurants, gyms, salons, and more. The data was the asset; what it needed was a public platform that could turn it into something people actually use: fast pages, real search, and price history, all holding up under heavy traffic and discoverable across millions of URLs. We built and run that platform — not the data behind it.
- Present a massive, constantly-growing dataset as a fast public product
- Make tens of millions of prices searchable, not just stored
- Show price changes and averages, not just today’s number
- Serve millions of public pages quickly, under real traffic
- Win search at that scale — every brand, item, and location
-
II Build What we built
pricelisto.com is the consumer-facing platform on top of the client’s price data. We structured how that data is presented and shipped the public product around it — brand and location menu pages, full search, price-trend views, and a multi-subdomain setup that splits the site by brand and by category.
- Brand and location menu pages with current prices, averages, and bestsellers
- Price-change history and trend views built from the underlying data
- Full search across items, businesses, and locations
- Calorie and nutrition pages linked to the same canonical items
- Multi-subdomain architecture — brand, category, and regional sites
- Admin tooling so editors can curate and correct what’s published
-
III Highlight Built to stay fast at 70M rows
The hard part isn’t storing the data — it’s serving it fast. At this size, naive queries fall over, so almost nothing is counted live: headline totals, averages, and popularity rankings are precomputed and cached, price history is served by unioning live and snapshot tables in a single query, and pages are cached whole and in fragments behind a CDN. Reads run off a database replica through a persistent application server, so the site stays quick no matter the load.
- Precomputed, cached aggregates and weighted ranking — no live COUNTs
- Live and historical prices merged via snapshot tables for trend pages
- Full-page and fragment caching with Cloudflare edge purging
- Read-replica split and Laravel Octane / RoadRunner for throughput
- Chunked background jobs, so heavy work never blocks a page load
-
IV Reach Found at scale
A platform like this only works if the pages get found. With millions of URLs, that means generating and refreshing sitemaps in cached chunks, keeping structured data on every page, and preserving search equity through restructures — so every brand, item, and location is its own fast, indexable page.
- Chunked, day-cached XML sitemaps covering millions of pages
- Structured data (schema.org) across brand, item, and location pages
- Clean, country-scoped URLs and a maintained 301 map for SEO equity
- Internationalized pages across dozens of locales
- Stack
- Menu-price platform · Search at scale · Price-change history
- Year shipped
- 2023–2026
- Status
- Live · pricelisto.com
- Filed by
- Dev Paragon · 2023–2026