What to actually consider when repricing your singles

Most card shops end up with one of two pricing problems. Either prices are stale — yesterday's market on today's shelf, which means chase singles under-price and reprint casualties over-price. Or prices are aggressive-against-market everywhere, which means the shop is permanently the cheapest seller and is quietly burning margin across every SKU. Both problems come from the same root: repricing rules set once years ago and never revisited.

Good repricing isn't 'match TCGplayer Low' on a cron. Good repricing is a per-segment strategy that accounts for market data source, condition, rarity, set age, channel-specific margin economics, reprint risk, and demand velocity. Here's how to actually think about it.

1. Pick the right market-data source for the segment

TCGplayer Market price (a weighted average of recent sales) is the right baseline for most US-facing retail. TCGplayer Low is a race-to-the-bottom signal — price at TCGplayer Low and you've volunteered to be the cheapest seller on TCGplayer, which is a fine strategy if you're scaling volume but a terrible one if you're trying to hold margin.

Cardmarket Trend is similar to TCGplayer Market but for European markets. If you sell in the EU, don't price EU inventory against US market data — the prices diverge, especially on foils and reprints.

CardSynced aggregates across regional markets. Useful if your customer base spans countries and you want a blended view.

eBay sold-listings matter most for graded cards and vintage. Sealed product 2020+ tracks eBay sold price more closely than TCGplayer Market because eBay is the dominant secondary marketplace for those segments.

Rule of thumb: modern retail singles → TCGplayer Market or Cardmarket Trend. Graded + vintage → eBay sold listings. Sealed → eBay sold + CardSynced.

2. Set different margins per rarity and per condition

Commons and uncommons live on volume. A 20% margin at 50 sales a day beats a 40% margin at 5 sales a week. Price commons aggressively vs market — market-minus-5% is defensible.

Rares and mythics have thinner supply curves — demand is less elastic. Market price or slightly below (market-minus-2%) is where most shops sit. Going further below market signals your whole shop as cheap and attacks your own rare-singles margin with no corresponding volume gain.

Condition multipliers matter more than most shops realise. If TCGplayer Market is the NM price, LP sells at 85-90% of that, MP at 60-75%, HP at 40-55%, Damaged at 20-35%. Don't use one global multiplier — Reserved List MP is different from bulk MP.

3. Price differently per channel

TCGplayer takes ~10.25% of each sale (8.95% commission + transaction fees). If your retail margin on a £20 card is 25%, selling at list price on TCGplayer leaves you ~15% margin after the marketplace cut. That's usually fine — TCGplayer volume is large. But you must price above your break-even, not at your storefront price.

Cardmarket takes a similar cut with different fee structure — model per-country because Cardmarket buyers compare prices by region.

eBay's final-value fee + payment processing totals ~12-14% on cards. Price accordingly.

Your own storefront has no marketplace cut. You can and should price slightly below your marketplace listings on your own site — 3-5% cheaper is enough to make direct-to-customer the obvious choice without racing the marketplace to the bottom.

The general pattern: direct storefront cheapest → Cardmarket → TCGplayer → eBay most expensive. Each channel's price reflects its fee structure plus a small margin buffer.

4. Set floors and ceilings

A floor price prevents autopricing from sending a £2 card to £0.05 because a bulk seller dumped a hundred copies on TCGplayer overnight. Common floor rule: minimum of market-minus-15% or £0.30, whichever is higher. Never price below cost — seems obvious, but autopricing without a floor does it constantly on commons.

A ceiling price prevents the opposite — autopricing a £10 card to £50 because TCGplayer Market briefly spiked on a single high-priced listing. Cap at market-plus-20% and manually review anything that hits the ceiling.

The floors and ceilings are your safety net. When they fire, your pricing engine has identified an edge case that needs human judgment. Review the audit log weekly.

5. Account for reprint risk

The biggest price movement in Magic isn't secondary-market drift — it's reprint announcements. A £40 card can drop to £8 in 24 hours when WotC confirms a Standard-legal reprint. Pokemon works similarly around Pokemon Center exclusives and set reprints.

Reprint-sensitive inventory needs tighter autopricing cycles (hourly, not daily) and active human review around spoiler season and Commander product announcements. If your autopricer reprices once a day, a Thursday-evening reprint announcement costs you 14 hours of margin bleed.

Some shops run 'reprint watch' lists — cards where a reprint is plausible (Reserved List-adjacent, Commander staples, format-defining cards). Those SKUs get more aggressive downward pricing rules and tighter refresh cycles.

6. Demand velocity — the hidden variable

A card with 10 copies sold per week on TCGplayer and a card with 10 copies sold per year should not be priced by the same rule. High-velocity SKUs can be priced aggressively vs market because you'll make it up on volume. Low-velocity SKUs should hold margin — if you only sell one per quarter, the margin on that single sale is all you get.

Most pricing engines expose a velocity signal (TCGplayer listing count, sales frequency, days-on-market). Use it as a rule input, not just a report.

7. The buylist price must move with the retail price

This is the one most shops get wrong. Retail repricing runs daily; buylist repricing runs weekly or never. Result: you overpay on buylist when the market drops, and underpay when it rises.

Rule: buylist rates are a percentage of market data, not a percentage of your own retail price. Pay 40% of TCGplayer Market for commons, 60% for rares, 85% for mythics. When the market moves, both retail and buylist move automatically. No manual intervention, no stale quotes.

8. Audit log and human review

Autopricing without an audit log is pricing without a brain. Every price change needs a before/after value and the rule that triggered it, kept for at least 90 days.

Weekly review: any SKU that moved more than 25% in either direction in a day, any SKU that hit a floor or ceiling, and any SKU with zero sales in 60 days despite being priced competitively. That last one is usually a condition-grade mismatch or a bad photo, not a pricing problem — but you only see it if you review.

9. Manual-override whitelist

Some SKUs shouldn't be autopriced. Artist signed copies, unique graded cards, anything with a handwritten note in the listing description, and anything a staff member has manually priced for a specific reason.

Keep a small whitelist of 'never touch automatically' SKUs and make sure your pricing engine respects it. A good rule: if a staff member has manually adjusted the price in the last 14 days, autopricing skips it for another 14 days.

10. The setup cost is paid once; the margin compounds forever

Good autopricing rules take 2-3 weeks of iteration to dial in. Most shops underestimate this and either (a) leave default rules on forever, or (b) give up after one bad week and go back to manual repricing.

Both outcomes cost real money. The shops that hold margin are the ones that iterate on their pricing rules for a full quarter — reviewing audit logs, adjusting per-segment rules, tightening floors and ceilings — and then let the engine do its work. The iteration cost is one-time. The margin improvement compounds across every sale for years.