Edge cases, single-game LGS vs full-multi-game retailer, what changes?

The build differs meaningfully at the edges. Two ends of the spectrum:

Single-game LGS (e.g. Magic-only or Pokemon-only):

• Singles catalog typically 5k, 25k SKUs, manageable with native Magento indexers.
• One game taxonomy (set + rarity + foil + condition), simpler attribute set, less PIM overhead.
• Tournament + prize-payout flow is the high-leverage feature, FNM regulars are the revenue base. Loyalty + store-credit accumulation patterns dominate.
• TCGPlayer integration is essential (singles auto-priced against Market). Cardmarket if you serve EU.
• Drop protection less critical, chase cards happen but are rare. Standard Turnstile + stock reservations are usually enough.
• Build cost: $5k, $12k. Timeline: 4-8 weeks.

Full multi-game retailer (Pokemon + MTG + Yu-Gi-Oh + sports + Lorcana + One Piece):

• Singles catalog 50k, 200k+ SKUs, needs custom denormalised stock-status table + heavy index optimisation + Hyvä storefront for category-page performance.
• Multiple game taxonomies in parallel, PIM-style attribute inheritance is the high-leverage architectural choice.
• Drop protection is critical, sealed Pokemon Booster Box drops, Black Lotus Beta restocks, graded Charizard inventory. Cloudflare Turnstile + WAF + per-customer guard + optional raffle mode all in scope.
• Multi-channel mandatory, TCGPlayer + Cardmarket + eBay + Whatnot + Discord. Channel manager (Codisto or ChannelAdvisor) at $400, $1k/mo.
• Tournament prize-payout matters less per-store but more in aggregate if you run satellite events.
• Build cost: $12k, $30k. Timeline: 10-20 weeks.

Most retailers I work with are somewhere in the middle, 2-3 games, 15k, 40k singles, primary channel TCGPlayer + own storefront, occasional chase-card drops. The free 30-min consultation maps your specific edge to the right build scope.