Sports-technology procurement often begins with a feature list and ends with an operating model that nobody fully priced. The questions below are designed to move the difficult conversation forward, before a contract turns assumptions into obligations.
1. Define the outcome before the product
What measurable sporting, commercial or operational outcome must the technology create? Which existing process will change if the implementation succeeds? Who inside the organisation owns that outcome after the sales and project teams leave?
A requirement such as ‘AI-enabled analysis’ is not an outcome. Faster review, additional sellable inventory, fewer on-site staff or a defined reduction in recovery time can be measured and governed.
2. Draw the complete service boundary
Which cameras, sensors, networks, power supplies, data feeds, venue interfaces and third-party approvals does the service depend on? Which of those are included in the supplier’s responsibility, and which remain with the rights holder, club, venue or broadcaster?
Who owns integration testing? A system can satisfy its own specification while the overall service fails at the boundary between two suppliers.
If a dependency is necessary for the service but absent from the responsibility matrix, it is already a future dispute.
3. Procure the failure path
What happens when the most advanced feature is unavailable? How quickly can operators identify the failure, move to a safe state and restore service? Is the fallback genuinely independent, or does it share the same network, timing source or control layer?
Ask for measured recovery time from a representative live test. ‘Redundant’ describes an architecture. It does not prove that people can recognise a fault and recover within the time available.
4. Make data rights explicit
Who owns raw data, derived data, annotations, model outputs and operational logs? Can the organisation export them in a usable format? May the supplier use competition data to train a model or improve products for other customers?
How long is data retained, in which jurisdictions, and what is deleted when the agreement ends? A right to access a dashboard is not the same as ownership or portability.
5. Test the commercial promise
If the system creates sponsorship inventory, how is delivery verified by market, feed and time window? If it reduces cost, which roles, travel, infrastructure or supplier fees actually disappear? If it creates efficiency, who receives the saved time?
Require a baseline and an agreed measurement method. Otherwise every positive result can be attributed to the technology and every shortfall to the operation around it.
6. Check the people model
How many trained operators are needed at venue and remotely? Who provides cover for illness, travel disruption and overlapping events? How long does competence take to build, and does the customer have the right to train its own staff?
Automation changes work rather than removing it. Procurement should identify who monitors the automation, who may override it and who carries accountability when the system is uncertain.
7. Put acceptance into the contract
Which tests must be passed before acceptance, and under what conditions? A laboratory demonstration, an empty-stadium test and a live match prove different things. Define the sample size, tolerances, failure severity and evidence required for sign-off.
Who decides that a defect is material, and what happens if acceptance is delayed? Payment milestones should follow proven capability rather than calendar dates alone.
8. Plan the exit before the launch
Can another supplier take over without rebuilding the entire stack? Who owns configuration, documentation, integrations and historical records? What assistance must the incumbent provide during transition?
A credible exit plan improves the current relationship. It prevents technical dependency from becoming commercial captivity and forces both sides to document the service properly.
The final three questions
Who has the authority to stop the service on match day? Who carries the financial consequence when it fails? What evidence will both sides examine on Monday morning?
If those questions do not have clear answers, the procurement is not finished, regardless of whether the contract is ready to sign.
Editorial note
This is a vendor-neutral editorial article. It does not describe a confidential tender, client engagement or named supplier.