Overview
How a Seed-Stage Team Built a Reliable Listings Dataset in 10 Days
Flowstack, a seed-stage B2B marketplace, needed a reliable competitor listings dataset before a pricing-strategy pitch to investors. They had twelve engineers and no one who wanted to own a scraper.
Starting point
A one-off Python script covered three competitor sites and broke every time one of them shipped a redesign. There was no schedule, no alerting, and no export format anyone downstream could rely on.
What they set up
- Discovery against each competitor's root URL—sitemaps and product pages found automatically
- Scheduled daily crawls with soft-block detection so a bot-served fallback page never got stored as real data
- Entity resolution matching the same product across all three competitors and their own catalog
- JSON export feeding straight into their internal pricing model
Results after 10 days
The dataset went from "whatever the script last managed to grab" to a daily-refreshed, deduplicated catalog. The CTO reports the pricing pitch would have used stale numbers without it.
Ready to try Scrapliy?
Turn competitor research into live catalogs, change alerts, and export-ready feeds.