Web-scraped alternative data went from edge to table stakes in under a decade. The problem has shifted from "should we use it" to "how do we source it reliably, defensibly, and before everyone else has the same series." Marquee Data has collected web data for hedge-fund clients since 2017; this is what that work has taught us.
The common thread: point-in-time collection, consistently, for long enough to build a history nobody can buy off the shelf. The value is in the time series, which means the value is in the collection continuing to work.
Most funds that scrape in-house discover the same thing: the quant who built the collector is now spending Fridays fixing it. Sources redesign, add protection, or rate-limit; a series with gaps is a damaged asset. A collection partner's job is for that never to be your problem — and for the fund's name never to appear in the traffic. The alternative, buying a standard dataset from a data vendor, solves reliability but surrenders exclusivity: a series sold to twenty funds is priced into the market. Custom collection is how a signal stays yours.
Alternative data desks now run real diligence, and web-scraped data has to survive it. What a defensible collection program looks like:
The US legal line of cases (hiQ v. LinkedIn through Meta v. Bright Data, 2024) has consistently distinguished public-page scraping from unauthorized access. A vendor should be able to explain exactly where their collection sits relative to that line — ours does, in writing, per engagement.
Marquee Data has built and managed custom web data collection since 2017 — Fortune 500 and hedge-fund clients, 100% US-based senior engineers, fixed fees. You don't pay until you're satisfied with the data.
Tell us about your project or email solutions@marqueedata.com.