Coupang Organic Rank Tracker — automated, ad-excluded search-rank collection

Coupang Organic Rank Tracker searches Coupang for the keywords you register and automatically collects and records only the ad-excluded organic search ranking of your product. Instead of typing into the search box by hand every day, you register your keywords and product once, then start, stop, and monitor progress and logs from a single web dashboard.

Why we built it

Coupang search results mix sponsored (ad) placements with organic ones. So when you just eyeball "what position are we in?", it is hard to separate the paid slots from your real search competitiveness. The organic rank that remains once ads are removed is the true result of your detail page, reviews, and sales signals — yet checking it by hand, across many keywords, every day, is both time-consuming and unreliable.

We needed a tool that records the ad-excluded organic rank only, across many keywords, continuously and automatically.

What it does

• Register keywords & product — just paste the search terms and the product URL. The product name is filled in automatically from the search results.
• Ad-excluded organic ranking — it actually searches each keyword, identifies and filters out sponsored (ad) cards, and computes the rank within the pure organic results only.
• Per-keyword rank table — for each product it lays out which keyword ranks where and when it was collected. If the product is not in the results, it is clearly marked as "not found".
• Live dashboard — progress, runtime, and average collection cycle update in real time, with a "system log" (remote control & errors) kept separate from the "activity log" (the collection process) so you always know what is happening.
• Continuous mode — start once and it cycles through the keyword set, recording rank changes over time.

The hard part: running for a long time without stopping

Automating the search itself is the easy part — running for a long time without getting blocked is much harder. Coupang blocks access when it detects automated traffic, so the tool waits a fixed cooldown based on the type of block and then resumes automatically with a fresh browser; if the browser dies unexpectedly, a watchdog detects it and relaunches. If it keeps failing at the same point, it stops safely instead of looping forever. It is designed to recover on its own during long, unattended runs.

Tech stack

• Web panel — a control panel (dashboard, rank check, logs, settings) built with Next.js and TypeScript.
• Collection engine — it drives a real Chrome browser over CDP, reading rankings from the same page a human would see.
• Local storage — keywords, products, rank snapshots, and logs are written atomically to local files, so history survives restarts.
• Testing — core logic such as rank calculation and ad detection is covered by roughly 200 automated tests.

Outcome

What used to be a manual search became automatic, letting us accumulate the "real" ranking — the one ads do not hide — on a per-keyword basis. It is now clear at a glance which keywords are on top and which have slipped, so the next move (detail-page improvements, keyword strategy) can be decided far faster.