Whatnot streaming for card dealers — the pre/show/post playbook

A Whatnot show moves cards faster than any other channel in the hobby. A two-hour stream can clear $3,000-$8,000 of inventory for a mid-volume Pokemon dealer. That's the upside.

The downside is that an unprepped show is worse than not streaming at all. You stumble through pricing, you forget to scan an SKU, you go live without restock cards, your average bid drops 30%, and the next show you announce has 40% fewer viewers because the algorithm noticed.

This is what I do for every Whatnot show. It works for me and it'll work for any dealer at any volume. The framework holds.

The 72-hour view

A Whatnot show is a three-act operation:

1. Pre-show (72 hours before to 1 hour before) — inventory selection, pricing, product list build, marketing, sleep.
2. Show (2-4 hours live) — execute, react, ship the audience to the next show.
3. Post-show (immediate to 48 hours after) — pack, ship, reconcile, schedule next show.

I'll walk each phase.

Pre-show — 72 hours out

Pick a theme

Whatnot's algorithm rewards specificity. A "Pokemon" show is too broad. A "Crown Zenith Galarian Gallery Singles" show pulls a tighter, more buyer-intent audience.

Themes I've run that worked:

• Crown Zenith singles only
• Japanese Pokemon vintage (1996-2003)
• "Under $25 chase" — every card has a comp above $25 but starts at $1
• Slabs only (PSA / CGC)
• Sealed product breaks

Themes that didn't work for me:

• Multi-set "mystery" boxes — too gambling-adjacent for the Pokemon audience
• Mixed sports/Pokemon — audiences don't overlap as much as you'd think
• Pure bulk — Whatnot's not the right channel for $0.50 commons

Pick a theme that lets you build a 90-180 card show with no filler.

Build the product list

Tag every show card in your inventory system with a show identifier (whatnot-show-2026-06-08 or similar). 120-150 cards minimum for a 2-hour show; 180+ for a 3-hour show. Plan for 60-90 cards/hour of live throughput.

Each card needs:

• A starting bid you're willing to actually accept
• A buy-it-now you expect 10-20% of cards to hit
• The right photo (front, ideally back too)
• A short title — Whatnot's UI cuts off long ones

Whatever inventory tool you use, build the show list as a batch you can export to Whatnot's product-list importer in one shot. Hand-uploading 150 cards through Whatnot's UI is a 90-minute slog you don't need. (Vault note: direct Whatnot integration is on our Phase 5 roadmap; until it ships, use CSV export from your inventory of choice.)

Set the starting bids

Two schools of thought on starting bids:

The dollar-floor school. Every card starts at $1. The bidding does the price discovery. Risk: a card you'd never sell raw for $40 sells at $4 because there are only two bidders in the room at that moment.

The 40% school. Every card starts at roughly 40-50% of the comp price. Risk: bidding momentum stalls because the starting bid felt high, and the card doesn't sell at all.

I run a hybrid: $1-start for cards under $20 of comp value (where the BIN safety net kicks in), 40%-start for cards above $20. The exact split depends on the audience size you expect.

The buy-it-now safety net: every card has a BIN at roughly 1.0-1.1x comp. About 12-18% of my show cards sell BIN rather than going through bidding. The BIN is the protection against a card going for 10% of comp; the bidding is the upside on cards that draw interest.

Marketing the show

72 hours before show time:

• Post the show on Whatnot itself. The platform's discovery feed favors well-tagged shows.
• Post a teaser to your Twitter/X with 4-6 of the chase cards.
• Post to relevant Discord servers if rules allow.
• Email your buyer list if you have one (most dealers don't — start one).

12 hours before:

• Reminder post on socials with the "now in 12 hours" framing.
• Update your Whatnot show description with the final card-count.

1 hour before:

• Go live in the show's pre-roll briefly to test mic/camera. Whatnot's audience trusts shows that don't have technical problems.

Sleep

If your show is at 8 PM, don't be running through last-minute card pulls at 7:55 PM. Have the inventory pulled, sleeved, and stacked next to you by 5 PM. The hour before showtime is for snacks, water, and reviewing the runsheet.

During the show

The opening

The first 8-10 minutes of any show are the lowest-attendance and lowest-velocity. Use them well:

• Don't lead with chase cards. Save them for the middle when the room is full.
• Lead with $5-15 cards at $1-start. These move fast, build bidding momentum, and signal to the algorithm that the show is "hot." Whatnot's discovery surfaces hot shows.
• Greet new viewers by name when they join. The first 50 people through the door are usually repeat buyers from prior shows; they care about being seen.

Mid-show

By 30-40 minutes in, the audience is built. This is when you run the chase cards:

• Slabs first if they're the headliners. A PSA 10 Charizard at $1-start commands attention.
• High-end raw next.
• Mid-range chase last.

Pace check: you should be selling 1 card every 50-70 seconds for a 2-hour show with 120 cards. If you're slower, the audience drifts. If you're faster, you're not letting the bidding develop.

The runsheet (and why it matters)

I keep a paper runsheet on my desk. It has:

• Card number (matches the SKU on the back of the physical card)
• Title
• Starting bid
• BIN
• Comp (for me to verify the closing price isn't catastrophically low)
• Channel status (already cross-listed on eBay? if so, pull it before going live)

The runsheet is my source of truth. I do not trust Whatnot's UI to show me the next card in the queue — sometimes it lags, sometimes the queue order changes when a buyer asks me to "show that next card again." The paper runsheet means I can't lose my place.

Pulling cross-listings

The biggest operational risk during a show is selling a card on Whatnot that's also live on eBay or TCGPlayer. If both sell, you refund one buyer and take a reputation hit on the channel you refund.

The right pattern: 5 minutes before show time, end the eBay listings for every card going into the show. After the show, re-list the un-sold cards. Painful manually; automatable if your inventory tool supports it.

We're building this into the vault — a "show mode" toggle that ends eBay listings before the stream and restores them after, with Whatnot integration to follow when the broader Phase 5 channel work lands. Until then: the manual pattern works, and it's better than the alternative (one double-sold card costs hours of refund-and-apology).

The buyer questions

Two questions every Pokemon Whatnot audience asks:

1. "Is that the holo / non-holo?"
2. "What's the comp on this?"

Have both answers ready for every card. The runsheet has the holo/non-holo column. The comp column is what I tell the audience when bidding stalls.

Telling the audience the comp is a power move. It justifies the starting bid, it signals you've done your homework, and it sets the buyer's expectation for where the card "should" close. I do this on every card above $25.

Post-show — the next 48 hours

The same night

When the stream ends, do not log off and go to bed. Spend 30 minutes:

1. Re-list un-sold show cards back to their original channels.
2. Verify the sold cards are flagged as sold in inventory.
3. Print labels for sold cards (the vault batches this).
4. Pull each sold card from its location, sleeve it, tape it into a top-loader, and stack it on the pack-out table.

If you skip this, you wake up to chaos.

The morning after — pack-out

Whatnot's shipping window is short. The buyer expects shipment confirmation within 24-48 hours.

A 100-card sold haul takes me about 2-2.5 hours to pack:

• Sleeve + top-loader: 30 seconds per card.
• PWE or bubble mailer envelope: 20 seconds per card.
• Tape and label: 20 seconds per card.

I batch by destination. All bubble mailers first, all PWEs after. The vault prints labels in the same batch order as the inventory location string, so I'm walking the pack-out room in one efficient loop.

24 hours after — reconcile

Inside the vault, I reconcile:

• Whatnot sales vs my "sold" count from the show. They should match.
• Whatnot's commission (8% as of mid-2026) calculated against my gross. The vault tracks this; you'll need to too if you want honest margin math.
• Shipping fees collected vs shipping fees paid. Whatnot's "buyer pays shipping" model is buyer-friendly but I lose roughly 6-8% of gross to overage shipping on heavier packages (slabs, multi-card lots).
• Refund / dispute pile. Whatnot's dispute rate is roughly 1.5-3% of orders for me; budget for it.

48 hours after — schedule the next show

Whatnot's algorithm rewards consistency. A weekly cadence on the same day of the week outperforms sporadic shows. I run Tuesdays 8-10 PM ET, every week, no exceptions. The audience trains itself.

Pick a day. Announce the next show within 48 hours of the current show ending. The descending viewer count from your current show flows into the announce-impressions for the next.

What I won't do on Whatnot

• Sell bulk under $5. The margin on Whatnot's 8% commission + shipping subsidy doesn't pencil.
• Run "mystery" anything. Audience is more gambling-fatigued than they were 18 months ago, and Whatnot has been tightening rules.
• Show cards I haven't physically pulled. "I'll grab it from the back" mid-show is a dead show.
• Stream for more than 3 hours. Audience exhaustion compounds fast. Better to run two 90-minute shows in a week than one 4-hour show.

My current numbers

Per show, on average:

• 120-150 cards in the product list
• 2-2.5 hours live
• $3,800-$5,400 gross
• 96-105 cards sold
• 1.5-3% post-show dispute rate
• ~$50-90 in shipping overage
• 6-8% audience growth show-to-show on the current cadence

That's the math. Whatnot is the highest-velocity channel I run and it's worth every minute of pre-show prep.

— Jamie

Whatnot is on the vault's Phase 5 channel roadmap. When it ships — direct product-list export, cross-listing pull during showtime, post-show sold sync — it lands on every tier with no upgrade gate. See What's coming next in the vault for the build queue.