From February to May I sold Pokemon cards every single day for 100 days and recorded everything. The totals: $7,450 in sales across 1,170 orders, $1,229 in fees, $6,192 paid out, average order $6.37, at roughly a 20% margin on inventory bought at bulk rates and ~60% of market. I started the run with a $2,000 collection buy and ended with about 53,000 cards in inventory. Here are the five lessons that actually mattered, with the numbers attached.
1. Direct carried the volume
TCGplayer Direct stayed the best way to sell: TCGplayer fulfills the orders, and I shipped seven consolidated reimbursement invoices instead of hundreds of envelopes. At a $6.37 average order, per-order packing would have eaten the margin whole. Penny-and-a-half trainers selling at market through someone else's fulfillment is as close to leverage as this business gets.
2. Free shipping over $5 is a volume lever
Mid-run, sales velocity dipped as my hottest bulk sold out, so I switched from priced shipping to free-over-$5. Two effects, both real: listings jump to the top of TCGplayer's price-sorted results (the platform sorts on all-in price), and buyers pad carts toward the free threshold, raising order value. Below $5 the buyer pays a dollar shipping anyway, which covers the envelope. Priced shipping earned more per order earlier in the run; free shipping moved more cards later. Match the lever to the inventory.
3. Competitive cards are the engine
Outside Direct, most orders were tournament staples: the format's trainers and techs, not chase Charizards. My competitive background (I played to top-16 North America) is why I stockpiled the right trainers ahead of rotation, and that stash sold all run long. The repeatable edge: acquire competitive cards at bulk rates, sell at market, and watch the format, not the hype. Which cards those are is a live-data question: the worth-money guide covers the framework and every card's current price is on its page here.
4. The fees are a fifth of everything
$1,229 of $7,450 went to fees: about 16.5% before shipping supplies. That is not a complaint, it is a budget line, and it is why the cost-model discipline matters: at a $6 average order, a seller who has not modeled fees to the cent is guessing whether they are profitable, and usually guessing high.
5. Consistency was the actual product
Eleven-plus orders a day for a hundred straight days is an operations claim, not a card-knowledge claim. The systems did the work: sorted inventory so every pick took seconds, shipping methods priced to the penny, and buy rates that kept inventory flowing in at margin-positive prices. The 100 days proved the business model; it also proved the model is a job. Whether that trade appeals to you is the honest question this page cannot answer.