The CardBot Killer? Our Take On The Roca Sifter

After 6 years of attending GAMA, this was certainly the most exciting and eventful one for CardCastle. I personally had an excellent time reconnecting with everyone in the industry.

If you're at all involved in TCG selling, eBay's marketing firehose has already made you aware of their new $800 sifting machine.

Naturally our customers have been asking what this means for CardCastle and the CardBot, so here are the insights which emerged from discussions with operators game store owners, major TCG operations, as well as some of those involved in the rollout of the Roca Sifter.

Keep in mind that our understanding of the development, capabilities and performance of the Roca Sifter has been informed through conversations - and people say all kinds of things. The facts and information frequently conflicted and we had to stitch together the most logical narrative ourselves.

We'll just have to wait and find out when they start landing in September and getting tested in the wild.

🧠 What We Understand

Key Specs

• Price: $800 + $300/year = $1,100 first year cost - so cheap it's almost too good to be true!
• Speed: 1,800/hour. Running at only ~800 cards/hour during live demo
• Capacity: 400 cards. Seems too low at first glance, but our metrics show the average collection buy at the counter is around 300 cards
• Foil Detection: They use a metal detector, the same tech used to identify staples in pages. It will detect foil vs non-foil, but cannot distinguish finishes (e.g Reverse Holofoil)
• Handles Sleeves: Huge if true. While the live demo showed the ability to handle mixed sleeved and unsleeved cards, the website says they must be run in separate batches.

Development - An eBay Initiative

According to our sources, eBay has been working on this for at least 2 years. The contract originally went to Epson, who must have lost it for some reason because it was then handed to Lexmark.

Lexmark was acquired by Xerox, who is now licensing the hardware to eBay. It seems TCGplayer has been handed the project from up high and directed to build the software integration into their Scan & Identify solution.

Takeaway: This is as enterprise as it gets - multiple disparate R&D teams, IP licensing deals, proprietary technology, and several brand reputations on the line (there are at least 3 logos slapped onto the side of the new sifter)

Announcement & Pre-Orders

There was a lot left to be desired from the live demo at GAMA:

• Speed: A number of creators on YouTube have pointed out that the machine was running at less than half-speed (~800 cards/hour)
• Sifting Filters: Missing most of the critical ones, and the price sift was hard-coded at $0.30+
• Orientation: Despite the website stating the machine can identify cards in any orientation, the functionality is yet to be built out
• Data Management: The interface was missing functionality to distinguish the cards in each stack, sorting and filtering ability, and scan summaries. Not important for a quick sift, but unlocks a lot of inventory management solutions.

While the R&D team at TCGplayer is certainly capable and experienced, we got the sense that there are already cracks forming in the Roca Sifter project due to a combination of pressures from eBay corporate, the tight launch timeline, and the current mismatch between what's on the brochure and what was shown off at GAMA.

Regardless, the hype is real and sentiment is incredibly positive from the backpack seller market. We assume they're targeting ~1,000 pre-order units, but wouldn't be surprised if they've underestimated demand and will hit over 2,000 orders.

Will they cut it off and post up a big SOLD OUT banner? Only time will tell...

The Takeaway: We're operating on the assumption TCGplayer will nail the launch, but at this stage there are several ways this project could flop and lead to a significant backslash from TCGplayer sellers.

Platform Tie-In & Integrations

You must have a TCGplayer seller account to use the machine, and this is revealing. We assume there will be an option to export data to a standard CSV, but it's highly unlikely they will provide integrations with third-party POS or E-Commerce platforms.

Let's look at the potential revenue. Assuming 1,000 units for the pre-order, the Roca Sifter launch will pull in $800k in hardware sales, $300k ARR (subscription revenue), totaling $1.1M total revenue in the first year.

Takeaway: The Sifter will provide a nice revenue boost, but its a drop in the ocean when compared to TCGplayer's annual revenue from marketplace fees alone. This is a loss leader, and the real money will be made in the increased marketplace listings.

Data & Server Load

It's unclear whether card recognition is onboard or cloud-based. It's safe to assume that there's little to no compute power inside the $800 device, so we're leaning towards cloud-based recognition. If so, they'll be uploading high-quality 300dpi images for 1,000+ units. At 2MB per image, that's 10TB/month hitting TCGplayer's infrastructure.

Takeaway: The team at TCGplayer shouldn't be strangers to software at scale, but supporting the influx of photography-grade scans and running neural nets over millions of sifted cards will be a decent lift.

😬 Is CardCastle Worried?

We've been in this game for almost 12 years now, starting with mobile app scanning and player collection management in 2014, prototyping the CardBot in 2017 and launching at GAMA 2020.

Our products and services are built with local games stores at the center.

Here's our take:

Price Isn't Everything

Our best customers invest in solutions that streamline operations and make them more money. The average CardBot processes over $40,000 in gross merchandise value each and every month. A store averaging just a 10% margin on those sales pays off their CardBot in 2 months.

It's Not About The Hardware, It's About The Ecosystem

CardBot isn't just a machine, its the first part of an integrated inventory solution capable of supporting your entire TCG operation.

The Roca Sifter requires a TCGplayer seller account, and they're likely selling it at a loss. The hardware is a way to lock you deeper into their marketplace ecosystem.

We're building the opposite direction. For example, our partnership with ShadowPOS means 1-click sync from your CardBot to your point-of-sale system. Scan a collection buy with CardBot and send it straight through to ShadowPOS - no CSVs required. Sell a card at the counter, your online store, or through partnering marketplaces, and your inventory will update automatically.

This is what marketplace independence actually looks like - your inventory, your sales channels, your business.

Hardware Fails, Support Matters

We pick up the phone. We're available 18 hours, 6 days a week - and our average live chat response time is 6 minutes!

With each CardBot purchase, we provide multiple onboarding sessions for your staff. We ship replacements free of additional charges, and frequently replace units as part of our Hotswap Program - report a critical hardware failure and we'll respond with a tracking number for your replacement. The faulty CardBot goes in the box and sent back.

Compare this to 6-8 hours for Rocasorter support at best, where the reply is often a request to fill out a support ticket which gets lost in the pile.

Sentiment for TCGplayer is at an all-time low with no signs of improving any time soon. Increasing fees, lack of service innovation, delayed payments, and numerous shipping errors have left both buyer and sellers frustrated. Now they're pivoting into yet another product line and rolling out cheap hardware to thousands of amateur sellers who will expect the device to run precisely as advertised.

Competition Is Good For Business

TCGplayer has validated what we've been saying for years - automation is the future for TCG retail, and sifting is key.

We're excited TCGplayer is reinvesting here. It means more sellers are thinking about workflow optimization, which is what we've been building for.

If you're a backpack seller running out of your basement and fully dependent on TCGplayer, maybe the Roca Sifter is right for you.

If you're an LGS supporting a local community, diversifying sales across multiple channels to protect your margins, and want to work with technology made by people who care - the CardBot is the better path forward.

So we're not changing course. We're going to keep building for businesses who want:

• Independence from marketplace lock-in
• Actual support when they need it
• Tools that scale alongside their business

📆 The Next 6 Months

Over the next six months we do expect the Roca Sifter launch to impact our sales from the backpack seller side of the market, especially in the short-term ahead of the September delivery date.

When those units land some sellers will love them, but we expect plenty of sellers will realize they will need more. Many will have issues and discover TCGplayer's support isn't what they hoped.

In the meantime, we'll be here, building an alternative ecosystem with partners like ShadowPOS. We'll be working tirelessly to make running your game store easy and fun. The way it's supposed to be.

Are You Selling Beyond TCGplayer?

The Roca Sifter is a tool for TCGplayer sellers. If you're running your LGS with a combination of in-store, online and marketplace sales channels, you need a different solution.

The CardBot integrates with more than a single marketplace - it integrates with your entire operation. Book in a demo today with the CardCastle team and let's have a chat about your business.

Not ready for TCG automation, but held back by your Point of Sale? Have a chat with the guys at ShadowPOS!