Platform Lock-In in Unattended Retail: A Contract and Data-Portability Checklist

Key takeaways

• Lock-in hides in five layers: hardware, payments, data, workflows and contract terms. Check all five, not just the monthly price.
• In small-ticket retail the fixed fee matters most: on 1.15 million transactions a year, every $0.01 per transaction is $11,520.
• Ask what switching processors actually requires: new credentials, re-keyed readers or new hardware. That answer is your real exit cost.
• Portability you haven't tested is a promise. Export a full data set and reconcile it to settlement once a year.
• Write change-of-control and price-change exit rights into the contract. In a consolidating market, your vendor may not stay your vendor.

Nobody signs up for lock-in. It accumulates. A card reader that only talks to one processor. A sales history that exports as monthly summaries instead of transactions. A renewal window that closed while you were busy opening 30 new locations. Each choice was reasonable when you made it. Together they decide whether, five years from now, you can negotiate with your platform or only accept its terms.

This guide is for operators and facilities buyers choosing or renewing unattended retail software. It covers where lock-in hides, how to price your exit before you sign, and the contract and data-portability checklist we’d want in hand. It’s practical guidance from an operator’s point of view, not legal advice. Have counsel review any agreement before you sign it.

Why this matters more in 2026

The unattended retail stack has been consolidating. The largest recent deal, 365 Retail Markets’ acquisition of Cantaloupe (valued at roughly $848 million), put micro-market kiosks, cashless payments and telemetry under one owner. Other vendors have been widening their own bundles, adding coolers, apps and AI features around a payments core.

Consolidation isn’t bad on its own. A larger company can fund hardware development, support and security work that smaller vendors struggle to pay for, and plenty of operators are well served by one bill and one support line. But consolidation changes the math of every contract. With fewer independent alternatives, switching costs carry more weight, and terms you accepted from one company may end up administered by another. The time to plan for that is during the negotiation, while you still have bargaining power.

The five layers where lock-in hides

Price is the layer everyone negotiates. It’s rarely the one that traps you.

Layer	How lock-in shows up	What to ask for
Hardware	Readers or telemetry modems that only work with one platform; proprietary firmware; no standard MDB or DEX support	Standard MDB and DEX/UCS interfaces; written confirmation that devices can be repointed or reused
Payments	Processing bundled into the reader or platform fee; a blended rate you can’t audit; settlement into the vendor’s account	Your own merchant account, or an itemized interchange-plus quote; the right to change processors without changing hardware
Data	Exports limited to summaries; extra fees for API access; access ending the day the contract does	Transaction-level export in documented open formats, any time, at no charge, plus a defined post-termination access window
Workflows	Planograms, pick lists, routes and driver habits that exist only inside one app	Exportable planograms, product catalogs and price history; an API for every object you create
Contract	Auto-renewal for a full new term; per-device minimums; termination fees that survive location losses	Short notice windows, renewal on monthly terms, exit rights on price changes and change of control

Two items deserve extra attention because they’re easy to miss.

Shopper-facing balances. If your micro markets run on prepaid accounts, the platform may hold your shoppers’ stored balances and loyalty points. Leaving then means refunding or migrating real money and asking hundreds of people to re-enroll. Find out who holds those funds, how they’re refunded, and whether account data can be exported for migration.

Employer programs. Subsidies and payroll deduct usually depend on an integration with the host’s HR or payroll system, and that integration may have taken the host’s IT team weeks to approve. If it only works with one platform, the host is locked in too, and your account renewal now depends on a vendor you can’t change.

Price your exit before you sign

Lock-in is a cost, so estimate it like one. Take an illustrative operator with 200 cashless-enabled machines averaging $1,200 a month in card and mobile sales, with an average ticket of $2.50.

• Annual cashless volume: 200 × $1,200 × 12 = $2.88 million
• Annual transactions: $2.88 million ÷ $2.50 = 1.152 million
• Every 0.25 percentage point of processing rate: $7,200 a year
• Every $0.01 of per-transaction fee: $11,520 a year

That last line is the one most quotes bury. In small-ticket retail the fixed per-transaction fee often matters more than the percentage, because a $0.05 fee is already 2% of a $2.50 vend before interchange is counted. If a bundled rate carries $0.03 more per transaction than you could negotiate on your own, the gap is $34,560 a year, or $172,800 over a five-year term.

Then add the one-time costs of leaving:

• Hardware swaps. If readers or modems must be replaced, multiply device count by the replacement price plus a site visit. Even at 15 minutes per machine, 200 machines is 50 technician-hours before drive time.
• Lost history. If you can’t take item-level history with you, any new forecasting system starts cold. That history is what lets a model tell real low demand from a slot that was simply empty, as our piece on forecasting beyond moving averages explains.
• Re-enrollment. Shopper accounts, badge mappings, payroll deduct authorizations and subsidy rules all need rebuilding.
• Termination fees. Whatever the contract says, often charged per device and per remaining month.

Write the total down. If the bundled offer still wins after you add exit costs, take it with eyes open. If it only wins because exit costs were left out, keep negotiating.

Processor independence: the clause that matters most

Of the five layers, payments is where lock-in compounds fastest, because the processor relationship can sit in three places at once: inside the reader, inside the platform and inside the settlement account. Ask three questions.

1. Whose merchant account is it? If settlement goes to the vendor’s account and they pay you out, you depend on their schedule, their reporting and their chargeback handling. If it’s your merchant account, you own the relationship and the data.
2. What does switching actually require? New credentials in a console is one answer. Re-keying every encrypting reader is another. Replacing the readers is a third. Encrypting (P2PE) readers are typically key-injected for a specific processor or P2PE solution, so ask before you buy 200 of them.
3. Is the rate itemized? Interchange-plus pricing shows what the card networks charge and what the processor adds. A flat blended rate can be fair, but you can’t compare it across vendors, and you can’t tell when interchange changes and the blend doesn’t.

Processor independence isn’t about distrusting any one processor. It’s about keeping one of your largest variable costs open to competition for the life of the contract.

The contract checklist

Bring this list to the negotiation, and let counsel turn it into contract language.

Term and renewal

• The initial term matches your confidence in the platform, not the vendor’s discount ladder
• Auto-renewal converts to month-to-month, not another full term
• The renewal notice window is reasonable, with no narrow window that closes months before the end date
• Annual price increases are capped and require written notice

Exit rights

• You can terminate without penalty if the vendor is acquired or materially changes pricing, features or support
• Early-termination fees are defined, capped, and exclude devices removed because a location closed or ended its agreement
• The vendor provides reasonable transition assistance: data, configurations and cooperation with your next provider

Data

• You own transaction, product, location and shopper-account data; the vendor’s license is limited to providing the service
• Transaction-level export is available any time, in documented formats (CSV, Parquet or API), at no extra charge
• Planograms, price history, product catalogs and device configurations are exportable
• Access continues for a defined period after termination, followed by deletion on request with written confirmation

Payments and hardware

• Bring-your-own-processor is allowed without a markup or platform penalty
• Readers and telemetry devices use standard MDB and DEX/UCS interfaces
• Hardware you buy stays usable, or can be repointed, if you change platforms
• Stored shopper balances are held and refundable under clear terms

Service

• The uptime SLA carries service credits, and its definition of “down” includes payments
• Support hours match your operating hours, because unattended retail runs at 3 a.m.

Run an exit drill once a year

Portability you haven’t tested is a promise, not a capability. Once a year, run the drill:

1. Export 90 days of transaction-level data, every field.
2. Load it into a spreadsheet or warehouse your team controls.
3. Reconcile totals against processor settlement deposits and DEX audit data. Any gap bigger than rounding needs an explanation.
4. Confirm that product IDs map to your own SKU list, not just the vendor’s internal codes.
5. Export planograms and price history, and check that someone outside the platform could read them.
6. Time the whole exercise. If it takes a week of support tickets, that’s your real exit speed.

The drill pays for itself even if you never leave. Several shrink categories, from restock mistakes to declined payments, show up first as a mismatch between what the platform reports and what the bank deposited. Our shrink taxonomy walks through each one.

When a bundle is the right call

Openness has costs. Running your own merchant account is one more relationship to manage. Mixing hardware brands means more spare parts and more training. A 15-machine operator with one route may reasonably prefer a single vendor, a single bill and a single support number.

So the test isn’t “bundle or open”. It’s whether the deal comes with a written way out. A bundled platform with month-to-month renewal, transaction-level export and bring-your-own-processor rights is a sound choice. An “open” platform with a five-year auto-renewing term and a summary-only export isn’t open in any way that matters.

How 8Kava approaches this

We built Kava OS on the assumption that customers should stay because it works, not because leaving is expensive. It works with MDB-compliant machines and readers and reads DEX/UCS audit data, with Kava Link connecting legacy machines, so existing fleets join without new cabinets. Data is available any time through an open REST API, webhooks and CSV or Parquet export. Payments are bring-your-own-processor at no markup, or Kava Pay at interchange-plus 0.35% + $0.05, and monthly plans carry no long-term contract and no setup fee.

For the full side-by-side, see Why 8Kava. Export formats and connectors are on Integrations & API, and plan terms are on pricing.