For brands
Advertise on the door your shopper is about to open.
8Kava's vending retail media puts your campaign on the 10.1-inch door display of a smart cooler and the 21.5-inch touchscreen of a smart vending machine, then records purchases on the same device. Run sampling, pilot a launch in a handful of buildings or open an unattended storefront, and judge it on closed-loop sales instead of reach estimates.
- connected retail points
- 3,400+connected retail points
- transactions in 12 months
- 11.8Mtransactions in 12 months
- checkout accuracy
- 99.4%checkout accuracy
Four ways brands use 8Kava
Retail media
Creative on cooler doors and vending touchscreens, scheduled by site type, daypart and audience. Impressions and purchases come from the same device.
Sampling
Trial units placed in the cooler and priced at zero or a discount for the campaign, with every unit that leaves accounted for by vision and weight.
Launch pilots
Test a new SKU in hospitals at night, campuses during finals or gyms at dawn before committing to a national reset.
Sampling with a control group, not a guess.
Street-team sampling hands out units and hopes. A cooler sample is measurable from the first unit. Merchandisers place your trial product in an agreed planogram slot, Kava OS prices it at zero or a discount for the campaign window, and vision and per-shelf weight sensing record each unit that leaves. Matched coolers that don't carry the sample become your holdout group.
When the campaign ends, you compare sites with the sample against sites without it: trial, follow-on purchases of the full-size product, and what shoppers bought instead. Analytics are anonymous and aggregated; recognition runs on the device, and no images of faces are stored or sent.
- Holdout sites matched by site type, size and daypart.
- Unit-level records from vision and weight fusion.
- Results by site type: hospital, campus, workplace, gym and more.
Our brand ran a sampling campaign across 60 coolers and could see which screens actually drove purchases.
How a closed-loop campaign runs
Choose the audience
Pick site types and dayparts: night-shift hospital staff, students after midnight, warehouse crews on third shift, commuters at dawn.
Set the holdouts
We match campaign sites with similar sites that won't run it, so lift is measured, not modeled.
Load creative and product
Creative goes to door displays and touchscreens, and merchandisers place product in the agreed slots.
Follow it live
Your team gets a read-only brand-partner view in Kava OS, with impressions and sales per device as they happen.
Get the result
Lift against holdouts by site type and daypart, exportable as CSV or Parquet for your own analysis.
Cooler screens vs. typical in-store screens
| Typical in-store screen | 8Kava door display or touchscreen | |
|---|---|---|
| Where it sits | An aisle, endcap or checkout lane | On the door of the product itself |
| What's measured | Modeled exposure, matched to sales later | Impression and purchase on the same device |
| Audience | Everyone who walks the store | A known site type and daypart |
| Smallest useful test | Often a chain-level commitment | A few dozen devices |
| Location host | Varies by retailer | Earns a revenue share |
Our guide to retail media on vending screens covers formats, pricing models and measurement in more depth.
Audiences you can reach
Healthcare
Clinical staff on nights and weekends, when the cafeteria is closed.
Higher Education
Students in residence halls and libraries, late at night.
Workplaces
Office employees on subsidized in-office days.
Fitness & Recreation
Members in the minutes after a workout.
Manufacturing & Logistics
Shift workers on short, fixed breaks.
Transit, Airports & EV
Drivers and travelers with a fixed wait.
What would you test if a shelf were this easy to rent?
A national launch commits you to slotting fees, a planogram reset and months of waiting for scan data. A pilot across a few dozen smart coolers shows trial and repeat by site type, with the same device recording what was shown and what was bought. Which product, price or message have you never tested because the test itself was too expensive?
Brand FAQ
How is the measurement closed-loop?
The door display or touchscreen that shows your creative belongs to the same device that records the purchase. Kava Vision builds each receipt from vision and per-shelf weight sensing, so impressions and sales line up per device and per daypart without third-party matching.
Do location hosts get paid?
Yes. Retail Media includes a revenue share for the location host on the devices where your campaign runs.
What screen formats can we use?
Kava Cooler has a 10.1-inch door-side display, and Kava Vend V4 has a 21.5-inch touchscreen. We share creative specifications when we scope the campaign.
Can we see results ourselves?
Yes. Kava OS has a read-only brand-partner role, so your team sees campaign impressions and sales without access to anything else on the host's or operator's account.
Is any shopper data personal?
Recognition runs on the device and looks at products, not people, and no images of faces are stored or sent. Reporting is anonymous and aggregated by device, site type and daypart. Read more on Security & Trust.
Where to go next
Bring us a product and a question.
Tell us what you want to learn: trial, price, placement or audience. We'll design the pilot, the holdouts and the readout.