How to Size Self-Service Dispensary Kiosks for Your Store

Use three inputs:

1. The number of orders your store completes during its busiest hour.
2. The percentage of peak-hour orders you want to move through self-service.
3. The number of completed orders one kiosk can handle during that hour in your store’s workflow.

Then calculate:

Peak self-service orders = peak-hour orders x target self-service share

Base kiosk count = peak self-service orders / completed orders per kiosk per peak hour

Round the result up to the next whole kiosk, then pressure-test that number against the floor plan, cash and payment mix, accessibility, network access, queue space, and fulfillment capacity.

That last step matters. The formula gives you a capacity estimate. It doesn’t tell you where the units should go or whether the rest of the operation can absorb the orders.

Daily averages hide the exact problem a kiosk is supposed to help.

A store may complete 160 orders across a ten-hour day and still do 35 of them in one sharp rush. Sizing from the daily average makes the floor look deceptively quiet when the line actually forms.

Pull transaction counts by hour from several normal weeks. Separate ordinary weekdays from your busiest days, and flag promotions, holidays, product drops, and local events that create unusual traffic. You’re looking for a repeatable peak, not the single busiest hour the store has ever seen.

Use completed transactions when you can. Door traffic is useful for staffing and conversion analysis, but kiosk capacity is tied to people who actually enter the ordering workflow.

For a new store with no transaction history, use the operating plan as a scenario, not a forecast you pretend is settled. Run a conservative case, an expected case, and a high-demand case. Build the floor so another unit can be added later if the demand shows up.

Not every customer will use a kiosk, and that’s fine.

The customer who knows the menu, buys the same products every week, or wants to browse privately is a natural fit for self-service. A first-time shopper who needs education may be better served by a budtender. The point is to give routine orders a direct lane, not force every customer into the same experience.

Start with what you can observe:

• How many customers arrive knowing what they want?
• How much of the store’s volume comes from repeat shoppers?
• How many customers already use online ordering or digital menus?
• Where does the staffed counter spend time on routine order entry?
• Will staff actively direct the right customers to self-service?

If the store has never used self-service, treat the adoption rate as an input to test. Don’t size the whole rollout around a benchmark borrowed from another retailer with a different menu, customer base, and checkout process.

This is where most kiosk ROI estimates get shaky. A vendor picks a transactions-per-hour number, applies it to every store, and produces a clean answer with more certainty than the data supports.

Session time changes with the job assigned to the kiosk. Browsing a large menu takes longer than checking a repeat order. Scanning information from an ID to pre-populate customer fields adds a different step than product browsing alone. Payment and receipt printing change the cycle again. Cash handling changes both the hardware and the customer flow.

Before a full rollout, run a pilot or a controlled floor test and record:

• Sessions started
• Orders submitted
• Orders completed
• Median session time
• Staff interventions
• Customers who leave the kiosk before submitting
• Payment method
• Time from order submission to fulfillment
• Unit downtime and the reason for it

Use completed orders per kiosk during a real peak hour as the denominator in the sizing formula. Median session time is helpful for diagnosing the experience, but completed orders tell you what the lane actually contributed.

A dispensary line can form in more than one place.

Customers can wait to browse, wait to enter an order, wait to pay, or wait for fulfillment. A self-service ordering kiosk adds capacity at the first two points. A payment-enabled or cash-enabled kiosk can carry more of the checkout path. Neither one fixes a fulfillment station that can’t pick and release orders fast enough.

Map the current customer flow from entry to exit:

Entry > check-in > browse > order > payment > fulfillment > exit

Mark every point where the customer stops and every point where an employee has to touch the transaction. Then decide which of those steps the kiosk should own.

If the store needs more ordering capacity, a standard self-service lane may solve the problem. If customers still have to rejoin a staffed line to pay, the operator should be honest about what that does to the queue. If cash is the constraint, review a cash-enabled dispensary kiosk as a separate lane rather than assuming every unit needs the same hardware.

The goal isn’t the largest kiosk count. It’s a clean handoff from the customer to the existing POS, payment, and fulfillment workflow.

Two stores with the same peak volume may need different hardware because the floors work differently.

The kiosk footprint is only part of the layout.

Customers need room to approach, browse, and leave without crossing the next person’s path. Staff need to see when somebody is stuck without hovering over every session. The store also needs an accessible route and enough clearance for the intended users, mounting height, peripherals, and local requirements.

Watch for the details that are easy to miss on a floor plan:

• A queue that blocks the entrance or pickup counter
• Screen glare at the busiest time of day
• A receipt printer staff can’t reach easily
• A cash unit with no practical service path
• A wall-mounted unit placed where browsing exposes customer information to the room
• Weak wireless coverage or no protected network path
• A kiosk that looks available even when the customer still needs to join another line

Tape the footprint on the floor before installation. Walk the route with two or three people using it at once. A placement problem is cheaper to fix with tape than after power, network, or mounting work is finished.

Kiosks can reduce the wait to browse and submit an order because they add ordering lanes. If fulfillment can’t keep pace, the visible line may simply move from the counter to pickup.

Compare peak-hour kiosk demand with:

• Orders the fulfillment team can pick and release per hour
• The number of active fulfillment stations
• Space for staged orders
• Staff coverage during rushes
• The process for exceptions, substitutions, and customer questions
• The handoff between the kiosk, POS, inventory, payment, and receipt flow

If the store can accept 20 additional digital orders but can only fulfill five of them, another kiosk isn’t the next purchase. The downstream process needs attention first.

This is also why two locations in the same company may need different kiosk counts. Standardize the method, hardware rules, reporting, and support model. Let each store’s actual demand determine the final unit mix.

The honest ROI model uses observed store data.

Track the total deployed cost, including hardware, configuration, payment or cash components, installation, network work, training, and support. Then measure what changes during the pilot:

• Completed self-service orders
• Staff minutes required per self-service order
• Staff minutes required for the comparable staffed order
• Peak queue length and wait time
• Order abandonment
• Fulfillment time
• Downtime
• Staff intervention rate

If you want to value staff capacity, use this equation:

Monthly staff capacity redirected =
(staff minutes per comparable counter order - staff minutes per kiosk order)
x completed kiosk orders per month / 60

That result is capacity, not automatic payroll savings. The hours may be redirected to fulfillment, education, exception handling, inventory work, or another part of the floor. Only count hard-dollar savings when the operating plan actually removes or avoids a cost.

Revenue claims need the same discipline. Don’t assign every kiosk order to “incremental revenue.” Compare completed transactions, abandonment, basket behavior, and peak-hour throughput before and after the pilot. If those numbers improve, use the measured difference. If they don’t, the kiosk may still have value in customer choice or staffing flexibility, but the ROI case should say that plainly.

A useful pilot has a question attached to it.

For example:

• Can one kiosk carry routine repeat orders during the Friday peak?
• Will customers use a wall-mounted ordering point near the entry?
• Does a cash-enabled lane keep cash customers from returning to the staffed counter?
• Can fulfillment absorb the added digital order volume?
• How often does staff need to step in?

Set the measurement plan before the unit goes live. Record the baseline from the staffed workflow, run the kiosk during comparable periods, and review what happened. The next unit should be justified by observed demand, not by the fact that the first unit looks busy in a photo.

For a new location, protect the option to expand. Put power and network where a second or third unit could go, leave space in the customer path, and make the initial hardware plan repeatable. You don’t have to buy every kiosk on day one to avoid an expensive retrofit later.