Vibe coding
Price tracking
Automation
Claude Code

Competitor price tracking on Amazon Mexico

Dreame · 2026 – Present

Did a competitor cut prices today? Instead of refreshing product pages by hand, let a system check every night.

Fig. 1. Competitor price tracking workflow

Company: Dreame
Marketplace: Amazon Mexico (MX)
Tracked: 32 ASINs · Dreame and Roborock
Stack: Claude Code · Bright Data · SQLite
Runs on: GitHub Actions · Netlify

01Context & challenge

Robot vacuums on Amazon Mexico are fiercely contested; competitor prices, Buy Box and stock change almost daily. Checking product pages one by one is slow and error-prone, and above all leaves no history, so pricing rhythms stay invisible.

The question: can we record price changes automatically every day, in a place colleagues can open any time?

02My role

I built the whole system, from fetching and storage to reporting and automated deployment:

Defined the watch list and fields: price, Buy Box seller, stock status
Wrote the fetching, parsing, storage and report scripts with Claude Code
Scheduled daily runs on GitHub Actions with automatic publishing to a fixed Netlify link
Wrote an operating manual so someone else, on another machine, can maintain it

03Approach

Fetch

Bright Data Web Unlocker fetches each product page; price comes from the page's structured data first, falling back to the buy box, with both first- and third-party seller layouts handled.

Store

Each day's data goes into SQLite, keeping the full price history, exportable as CSV for pivoting.

Report

A daily report with three views: today (rises, drops, stock-outs), price bands (where each brand concentrates) and trends (daily and monthly per product), with side-by-side comparison and a Chinese / English switch.

Automate

GitHub Actions runs it daily, commits the data back and publishes to Netlify. The link never changes, so colleagues bookmark it once.

04Outcome

Competitor ASINs tracked daily: 32
Runs and publishes automatically: Daily
Today · price bands · trends: 3 views

Competitor prices went from something checked when someone remembered to data that builds up every day:

Time saved: no more refreshing pages; one fixed link shows today's rises, drops and stock-outs
Visible trends: as history accumulates, competitors' pricing rhythm and price bands are plain to see, ready for pricing and promotion decisions