What Actually Changed at the Convenience Store in 2026
Forget the demos for a minute. What can we actually see?
There is never a shortage of new technology at NACS. AI, computer vision, autonomous checkout, RFID, robotics, smart shelves and whatever else happens to be fashionable that year.
That isn’t quite the question I am interested in.
I want to know what actually changed at the convenience store over the last 12 months. What can I walk into a store — or pull up to the forecourt — and actually see?
Using that test produces a somewhat different technology list.
The Forecourt Is Changing First
EV charging is probably the most obvious addition.
It is still a relatively small percentage of the enormous U.S. convenience-store universe, but where it is being deployed it changes more than the fuel mix. It changes dwell time.
A gasoline customer fills up and leaves. An EV customer may be there for 20 or 30 minutes. Suddenly foodservice, restrooms, seating, Wi-Fi and the overall quality of the store matter considerably more.
That helps explain another very visible change: newer convenience stores increasingly look less like gas stations with a store attached and more like restaurants that happen to sell gasoline.
Larger kitchens, made-to-order food, digital menu boards, pickup areas and self-order kiosks are becoming much more prominent.
QuikTrip’s new Generation 4 format is a good example. The kitchen is substantially larger, ordering kiosks get more space and there is a dedicated food pickup and waiting area. GPM’s newer Fas Craves format similarly emphasizes food, beverages, digital menu boards and customer flow.
That isn’t a technology pilot. You can walk into the building and see it.
The Gas Pump Is Becoming Another Digital Endpoint
The other change on the forecourt is the pump itself.
The fuel dispenser used to be a fairly simple payment terminal attached to a pump. Increasingly it is a digital display, payment device, loyalty terminal and advertising endpoint.
Weigel’s is a useful example. It is tying together fuel-dispenser screens, digital window signage and in-store displays across its locations.
That makes the forecourt part of the retailer’s media network.
It also creates an interesting measurement opportunity. If the retailer knows that I bought gasoline, saw an offer at the pump, entered the store and bought the promoted product, the pump display suddenly becomes measurable retail media rather than just another digital sign.
Circle K’s Full Circle Media initiative points in the same general direction.
Lottery Is More Important Than Most Technology People Realize
Lottery deserves much more attention in any discussion about c-store automation.
It is easy to install self-checkout in a grocery store and think the same model transfers directly to convenience. It doesn’t.
Convenience stores sell cigarettes, alcohol and lottery tickets. Those categories bring age verification, regulation and associate intervention into the transaction.
Lottery creates another complication. Move customers away from the cashier and you can inadvertently move them away from lottery.
That is one reason we are seeing more attention paid to self-service lottery, automated ticket inventory and tighter POS integration.
Scientific Games’ SCiQ technology is a good example of something actually being deployed. North Carolina has been expanding the technology across roughly 2,000 lottery retailers, and the platform has also appeared in retailers including 7-Eleven and Circle K.
The old lottery terminal and stacks of scratch tickets behind the cashier aren’t necessarily disappearing tomorrow, but the lottery transaction is becoming part of the broader self-service discussion.
Self-Checkout Is Real — But We Need Better Numbers
Self-checkout is also becoming much easier to find in convenience stores.
Smaller chains including Good 2 Go, Eagle Stop and MotoMart have been installing it, and foodservice operations inside travel centers are adopting self-order kiosks as well. Pilot, for example, has installed self-order kiosks across its Wendy’s locations.
The interesting part isn’t whether a kiosk can ring up a Coke and a bag of chips. We solved that problem years ago.
The real c-store question is what happens around the kiosk.
What happens with cash? Fuel? Loyalty? Tobacco? Alcohol? Lottery? Refunds? Coupons? An item the camera or scanner doesn’t recognize?
And what happens to the employee?
Those exceptions are where convenience-store automation either works operationally or becomes another device requiring an associate to babysit it.
Which Brings Us to Mashgin
Computer-vision checkout deserves attention, but this is where I would be particularly careful about separating announcements from installations.
Couche-Tard announced an intended rollout of Mashgin Smart Checkout to more than 7,000 stores. That is a very significant announced program.
It is not the same thing as independently verifying 7,000 operating installations.
That distinction matters.
Mashgin has real technology, real patents and real deployments. Circle K has publicly committed to the technology. But when we tried to document the visible U.S. convenience-store installed base store by store, the public evidence was considerably thinner than the headline rollout number might suggest.
That doesn’t mean the rollout isn’t happening. It means an announced target should remain an announced target until we can document the installations.
The same discipline should apply to everybody.
Electronic Shelf Labels Are Finally Showing Up
ESLs are another technology that has spent years being discussed in the United States and is now beginning to appear in actual convenience stores.
H&S Energy has begun deploying electronic shelf labels and expanding them across its estate.
The U.S. is still well behind some overseas markets. Asda, for example, has deployed roughly 1.2 million electronic shelf labels across all 517 of its Express convenience stores in the UK.
That gives us a useful picture of where this can eventually go.
The value isn’t merely changing a price without replacing a paper tag. Once the shelf becomes digital, pricing, promotions, inventory information and eventually retail media all start converging at the shelf edge.
RFID? Interesting, But Show Me the Store
RFID is a good example of why we need to separate technology possibility from operating reality.
There are good potential applications in convenience: receiving, reusable assets, higher-value merchandise, backroom inventory and selected categories.
But tagging every bottle of soda, candy bar and bag of chips is a different economic proposition.
I haven’t seen convincing evidence of a major U.S. convenience chain operating item-level RFID throughout the store at scale.
Until we do, RFID belongs on the radar rather than on the installed-deployment list.
And Then There Is Security
There is another technology change taking place that retailers would probably prefer customers didn’t notice.
Convenience stores and fuel stations remain attractive targets for payment fraud.
Skimmers are still being discovered inside fuel dispensers. Law enforcement continues conducting sweeps of gas stations looking for compromised pumps, registers, ATMs and payment devices.
The response is visible if you know what to look for: newer EMV payment terminals, contactless payment, tamper controls and upgraded dispenser electronics.
But the larger security problem isn’t visible.
A modern convenience store can now have POS terminals, self-checkout, food-ordering kiosks, fuel dispensers, EV chargers, lottery systems, digital signage, electronic shelf labels, loyalty applications, Wi-Fi, cameras and retail-media systems connected to networks.
Every new endpoint creates another operational dependency and potentially another attack surface.
The convenience store is becoming a fairly sophisticated distributed computing environment that also happens to sell coffee and gasoline.
What About AI?
AI is certainly being deployed, but much of the useful AI isn’t particularly visible.
Casey’s is using AI around forecasting and inventory planning. Voice ordering is expanding. Computer vision is being applied to checkout and loss prevention. Kitchen systems increasingly use data to determine what should be prepared and when.
Those may eventually produce larger operational gains than some of the shiny customer-facing technology.
But you can’t necessarily photograph them.
That is why I would separate visible store change from operational technology change.
For 2026, the things I can actually see are fairly straightforward:
EV charging. Bigger food operations. Digital menu boards. Ordering kiosks. Self-checkout. Smarter lottery merchandising. Bigger pump displays. Contactless payment. Some electronic shelf labels. More screens almost everywhere.
Behind those visible changes are the less visible systems — forecasting, computer vision, loyalty, retail-media measurement, cybersecurity and inventory intelligence.
And then there is the next wave: RFID everywhere, autonomous stores, frictionless checkout, digital identity and more sophisticated computer vision.
Some of those will undoubtedly become important.
But I would apply a very simple rule when walking NACS this year:
Don’t tell me what it can do. Tell me where it is running.
Better yet, give me the address.
Then I’ll go take a picture.
ADDENDUM – PATENTS
Yes. There is notable patent activity, but the interesting finding is that 7-Eleven looks quite different from Circle K/Couche-Tard when you examine who actually owns the IP.
7-Eleven is the one we will highlight
7-Eleven has a surprisingly coherent patent cluster around computer vision, customer tracking and cashierless/automated checkout. This isn’t just a generic “AI” patent.
A 2019-priority family assigned directly to 7-Eleven covers using video of a customer’s shopping session to create a virtual shopping cart. The family has resulted in U.S. and international grants, including a European grant in 2024. Google Patents
More interestingly, the surrounding 7-Eleven portfolio includes patents covering:
- customer-based video feeds;
- scalable position tracking inside large spaces;
- mapping sensors into a common coordinate system;
- object detection/exclusion zones;
- creation of a virtual representation of the physical store;
- automatically populating a customer’s virtual basket;
- and selectively verifying an algorithmically generated shopping cart before charging the customer. Google Patents
That last part catches our attention. The verification patent isn’t simply “camera sees Coke.” It deals with what happens when an algorithm has created the basket and you need a mechanism to verify it. That’s exactly the operational problem we were discussing with frictionless checkout: what happens when the computer is wrong?
There is also evidence the family continues to have value to 7-Eleven. A customer-video patent had its four-year maintenance fee paid in March 2026, and another member of the broader virtual-store family was granted as recently as August 2025. Google Patents
So we characterize this as a genuine strategic patent position, not a random filing.
Circle K / Couche-Tard is almost the inverse
We don’t see an equivalent body of retailer-owned Circle K or Alimentation Couche-Tard patents around checkout technology in this initial search.
That is actually interesting rather than disappointing.
Circle K’s highly publicized computer-vision checkout strategy depends heavily on supplier IP — particularly Mashgin. Your existing patent brief already caught this distinction: Couche-Tard announced a target of more than 7,000 stores, while the underlying object-recognition kiosk/visual-recognition patents belong to Mashgin. Pasted markdown
So you have two different strategic models:
7-Eleven: We own IP covering elements of the computer-vision store.
Circle K/Couche-Tard: We can deploy somebody else’s technology at potentially enormous scale.
Neither is inherently better. But it is a very useful distinction for your NACS research.
And given our earlier Mashgin discussion, I’d continue putting a large asterisk next to the 7,000-store figure: it is the announced rollout objective, not evidence that 7,000 locations are presently operating Mashgin.
Gilbarco may be even more interesting for your article
I’d expand the patent search beyond the retailers themselves.
Gilbarco has a 2023-priority international application titled “Systems and methods for an autonomous store.” It specifically discusses convenience stores and dynamically deploying store capabilities. The application was filed in April 2024 and published in October 2024. Google Patents
That fits your article beautifully because Gilbarco sits at the intersection of the forecourt and the store.
And its older portfolio shows how long this convergence has been developing. Gilbarco patents cover fuel-dispenser POS systems, mobile payment at the pump, secure payment interfaces and allowing customers to purchase convenience-store products or food from the dispenser. Google Patents
In other words, the idea we were discussing — the pump becoming a digital commerce endpoint rather than merely a fuel-payment device — has a substantial technology/IP history behind it.
Casey’s gives us another useful contrast
Oour existing research found something quite different there. Casey’s public strategy emphasizes AI forecasting, inventory planning, kitchen redesign, digital tools and prepared food, but oour initial scan did not establish a corresponding Casey’s-owned patent portfolio for those store systems.
That makes for an interesting three-way comparison:
| Retailer | What the patent trail suggests |
|---|---|
| 7-Eleven | Own IP around vision, customer tracking, virtual stores and algorithmic baskets |
| Circle K / Couche-Tard | Major deployment strategy using vendor-owned IP, especially Mashgin |
| Casey’s | Operational/data/food strategy appears more important than retailer-owned patents |
| Gilbarco | Supplier-side IP connecting forecourt, payments, POS and potentially autonomous stores |
We think that’s more interesting than simply compiling “c-store patents.”
The story becomes who owns the intelligence layer of the future convenience store?
7-Eleven apparently decided some of it was worth owning. Circle K is demonstrating that you can potentially get there through suppliers. Casey’s shows that proprietary operating data and execution can be more important than patents. And Gilbarco shows that some of the most consequential c-store IP may belong to the infrastructure vendors rather than the retailers.
We should add NCR Voyix, Verifone, Dover/Wayne, Invenco by GVR, PDI, VusionGroup, Zebra, Scientific Games/IGT, Diebold Nixdorf, Toshiba, Mashgin and possibly Amazon to the next patent pass. That’s where we’ll uncover the IP corresponding to the things we’re actually seeing in stores: pump media/payment, lottery, age verification, ESL, self-checkout, loss prevention and store/forecourt integration.