Wedding Pricing Compare

One-liner: A multi-source wedding vendor pricing research tool that scrapes DJ, coordinator, and photobooth pricing from TheKnot, Zola, and WeddingWire — surfacing market rates for a data-advantaged sales approach.

Origin (the Build Loop)

Because I sell wedding services and needed to know exactly what competitors charge across every metro, I kept hitting the fact that market-rate data exists — it’s just locked inside three giant marketplaces behind rate limits, bot detection, and GraphQL schemas, so I built a scrapers pipeline that collects pricing from TheKnot, Zola, and WeddingWire, deduplicates, and renders it as a clean dashboard, that lets me quote with actual market support instead of gut feel. It taught me that GraphQL introspection works for more than just APIs you own — TheKnot’s review graph was cracked open through standard introspection queries.

The problem

Wedding vendors need to know the market rate to price competitively, but the pricing data lives inside the three big wedding marketplaces (TheKnot, Zola, WeddingWire) — behind login walls, rate limits, GraphQL endpoints, and bot-detection. Individual vendors can estimate by calling around, but nobody has the actual distribution. I wanted the median, the range, the metro-level breakdown — and I wanted it from real listings, not surveys.

What it does

• Multi-source pricing collection — scrapes DJ, coordinator, and photobooth pricing from TheKnot, Zola, and WeddingWire.
• Deduplicated vendor database — cross-references vendors across sources using name, city, and service type matching.
• Market-rate dashboard — renders price distributions, medians, and ranges per service type and metro area.
• Rate-limit and bot-detection analysis — documented which sources allow what (TheKnot GraphQL is one of the most permissive; WeddingWire is the most aggressive with blocking).

How it’s built

• Stack: Python, Playwright, PostgreSQL, HTML/Chart.js dashboard
• Notable engineering:
  • TheKnot GraphQL deck — discovered via standard introspection on svc.theknotww.com/reviews-api/graphql. The x-tenant-id:tk-us header and storefrontId filter unlocked paginated review access. Comment.content and ratings[].value fields exposed full vendor feedback. This wasn’t a reverse-engineered private API — a standard introspection query showed the deck.
  • Playwright for Zola/WeddingWire — both sites require real-browser rendering; Zola is navigable with careful rate limiting, WeddingWire triggers blocking thresholds around 50 requests.
  • Cross-source deduplication — matching vendors across three sources with different naming conventions, same vendor in different cities.
  • PostgreSQL at scale — 1,734 vendors and 32,613 reviews across all three sources in a local PG instance.
• Architecture: 3 source adapters → normalize → PostgreSQL → Chart.js dashboard. Source adapters run independently so one source failure doesn’t block the others.

Proof points

• 1,734 vendors collected across TheKnot (1,593), Zola (91), WeddingWire (50).
• 32,613 reviews analyzed — TheKnot (31,128 via cracked GraphQL), Zola (1,485 via Playwright).
• Market-rate findings: DJ median $2,400, coordinator median $1,850 — these are real market numbers, not estimates.
• 985 vendors with reviews — enough for statistically meaningful per-category, per-metro distributions.
• 28 metro areas covered with DJ (787) and coordinator (836) as the deepest categories.

What to show

• Demo: Case-study-only (sensitive data — real vendor pricing). The dashboard showing price distributions per service type is the hero.
• Visuals needed: the price distribution histogram per service type; the metro-level median comparison; the source analysis table (permissions, rate limits, techniques); the GraphQL introspection cheat sheet.

Cross-links

• Shares the GraphQL introspection technique and the underlying vendor/review dataset with theknot-scraper, which analyzes the same data for sentiment/complaint patterns rather than pricing.
• The multi-source scraping discipline connects it to price-scrapers, EARLS, and infra-metals-dashboard.
• Feeds directly into SALES-BOT (market-rate anchoring for sales calls).

Case-study angle

This is data advantage in practice: a vendor who built his own pricing-intelligence pipeline because knowing the actual market rate is the single best sales lever. The GraphQL introspect-on-a-whim story is its own artifact — most developers don’t know that __schema meta-queries work on third-party APIs. It proves Alex finds data advantage where others don’t think to look (GraphQL introspection) and builds the pipeline to deliver it.