Web Scraping for E-commerce: Building a Daraz Price Tracker

Retailers across South Asia use Daraz price and stock data to drive their own pricing. Built well, a tracker runs on a small VPS and returns hours of manual work per day.

Legality First

Scrape only public product pages. Respect robots.txt where it applies to your use case, throttle politely, and never scrape private data or bypass paywalls. For commercial resale of the raw data, get legal advice in your jurisdiction.

Architecture

**Scheduler** — cron or Celery Beat, one job per category.

**Fetcher** — `httpx` with rotating residential proxies.

**Parser** — `selectolax` (much faster than BeautifulSoup).

**Store** — PostgreSQL with a `price_history` table.

**Alerts** — Telegram or email when a competitor drops price.

Fetching Pages

Use randomised delays and one concurrent request per proxy — anti-bot systems flag bursts, not volume. Rotate user agents and accept-language headers.

Parsing

Many e-commerce sites ship product data as JSON inside a