Skip to content

buy-pipeline.md

Latest commit

 

History

History
67 lines (53 loc) · 2.76 KB

buy-pipeline.md

File metadata and controls

67 lines (53 loc) · 2.76 KB
Raw

Buy Pipeline

The rare, deliberate exit from the funnel. Built LAST and kept minimal.

Principle

Purchase fires occasionally, not in the main loop. Do not over-build it. The vendors already do pricing, cheapest-seller selection, and shipping optimization — lean on them. The only custom value here is "buy only what I don't already own", which no vendor button does.

Two distinct shapes, only one needs the collection diff:

Shape A — Single card ("I want that card I saw in play")

No diff needed.

card name → TCGplayer Mass Entry URL → open browser tab

Bookmark-tier effort. Build this the first time you actually want a card.

Shape B — Whole deck ("I fell in love with this deck")

.dck file
  → parse to (qty, name) list
  → diff against local collection file (name + qty)
  → missing cards only
  → TCGplayer Mass Entry payload → open browser tab

Mass Entry then optimizes cheapest-per-seller across vendors and minimizes shipping server-side. Build this ONLY after first wanting a paper deck — let real usage tell you what it needs.

TCGplayer Mass Entry — the mechanism

  • It's the public web form their own "Buy Deck" buttons target — NOT the locked-down affiliate/developer API. No key needed.
  • Accepts newline-delimited quantity cardname lines, with or without set codes and collector numbers. Less specificity = "any printing".
  • Gear icon → preferences for acceptable printings/conditions (default: all printings, Moderately Played and better).
  • You construct the list and open the URL; the cart builds on their side.

Name normalization (the fiddly part)

  • Strip XMage set codes/collector numbers from .dck lines.
  • Handle split/DFC naming: TCGplayer may want the front face or full A // B depending on card. Test against a few real DFCs and adjust.
  • This is the main source of buy-step bugs; keep a small override map for the handful of cards that don't resolve cleanly.

Secondary vendors

  • Card Kingdom — has its own mass-entry; single-vendor, consistent condition, simpler shipping. Good for "what TCGplayer's cheapest sellers lack".
  • Cardmarket — relevant if buying from Europe.

Zero-code cross-check

Deckstats imports XMage decks, tracks a collection, and shows multi-vendor price comparison (TCGplayer, Card Kingdom, Cardhoarder, Miracle Games, Cardmarket). If you ever don't want to maintain the diff yourself, its collection tracker does the "what do I own" subtraction. Worth knowing as a fallback so you don't over-invest in custom buy tooling.

What NOT to build

  • No vendor-side cart logic — Mass Entry owns that.
  • No price scraping for the buy step — open the tab and let the vendor price it.
  • No speculative multi-vendor optimizer — Mass Entry already optimizes.