The Self-Checkout Text Nobody's Talking About

Self-checkout was supposed to be a labor play: fewer cashiers, faster lines, lower cost to serve. IHL Group's own shrink research shows it also opened a second, quieter cost. Stores with heavy self-checkout use see shrink rates 30 to 60 percent higher than staffed lanes, a gap the industry has started calling the "SCO tax." This week's Retail Reality Check breaks down Scandit's new camera-based Self-Checkout Loss Prevention tool, the IHL data on what AI-driven loss detection actually returns (374% ROI over three years, roughly $88,000 in prevented loss per store annually), and why three very different convenience-store chains are approaching the same rollout at three different speeds.

The back half is a fast-moving lightning round through 12 more stories: a $1.7 trillion global inventory distortion problem, Walmart's tap-to-pay rollout against Amazon's AI infrastructure bet, Instacart and Shipt's new AI shopping assistants, Coca-Cola's Perfect Basket AI in Malaysia, and a labor-scheduling AI catch that unlocked $20,000-$35,000 a month at a 10-store restaurant brand. Every number is sourced live in the episode or in IHL's published research, with no rounding and no invented quotes.

• Self-checkout stores carry shrink rates 30% to 60% higher than staffed lanes, according to IHL Group's shrink research, the basis for what retailers now call the "SCO tax"

• AI deployed specifically to catch scan avoidance and ticket switching at self-checkout delivers 374% ROI over three years and about $88,000 in prevented loss per store annually, with payback typically inside six months (IHL Group)

• Scandit's new Self-Checkout Loss Prevention tool uses cameras retailers already own to recover or deter more than 75% of self-checkout losses, per Scandit

• Diebold Nixdorf's kiosk modernization at Dave & Buster's hit 99.8% uptime while cutting wait times and raising throughput

• Williams-Sonoma's AI assistant Olive is driving 9x the revenue of an average visit (up from 2x a year earlier), and IHL's AI/ML research (16% higher sales growth, 134% higher profit growth in 2024) helps explain why that's realistic

• Retailers already using AI/ML in demand chain operations are 81 times more likely to have adopted headless commerce, the architecture behind Tapestry's new in-Gemini checkout (IHL Group)

• Inventory distortion, the combined cost of stockouts and overstocks, runs retailers $1.7 trillion a year worldwide, or 6.2% of global retail sales (2026 IHL Group research)

01:11 - Deep dive begins: checkout becomes the place retailers compete to see

01:44 - Scandit's Self-Checkout Loss Prevention and the 75%+ recovery claim

02:11 - The SCO tax: IHL's 30-60% shrink research and the 374% ROI figure

02:37 - Diebold Nixdorf modernizes kiosks at Dave & Buster's

03:01 - Eagle Stop, Good 2 Go and MotoMart roll out self-checkout

03:26 - Why MotoMart is capping its pilot at 10% of locations

03:47 - Williams-Sonoma's Olive and the nine-times revenue lift

04:12 - IHL's AI/ML sales and profit growth data

04:36 - Tapestry brings Coach and Kate Spade into Google Gemini

05:00 - Headless commerce and IHL's 81-times adoption stat

05:23 - Deep dive close: the technology already exists

05:48 - Lightning round begins: SKU proliferation and the $1.7 trillion inventory distortion problem

06:18 - Walmart's tap-to-pay rollout and Amazon's Qualcomm AI chip deal

06:45 - Instacart's Clementine and Shipt's Ask Shipt

07:14 - Coca-Cola's Perfect Basket AI in Malaysia

07:44 - Costco, Walmart and Wakefern expand delivery partnerships

08:10 - Skechers builds a new automated logistics center in Belgium

08:33 - Philippine Seven Corporation and SynergySuite's labor scheduling win