A manufacturer can restore its office email and still be unable to ship a product. Production may depend on scheduling systems, machine programs, industrial controllers, labeling equipment, and outside service connections. A cybersecurity plan needs to account for those dependencies.
When selecting a managed security provider, start with what the plant must keep running and how it would recover. A generic list of security tools is not enough to establish that fit.
Map the production dependencies
Identify critical systems, their owners, and the vendors allowed to access them. Include older equipment and systems that cannot tolerate ordinary office-IT maintenance practices. Document which processes depend on a shared network, server, identity system, or cloud service.
NIST’s manufacturing cybersecurity resources provide guidance designed for manufacturing environments. Use such frameworks to organize the discussion, then require the provider to explain how its proposal fits your actual plant.
Ask how the provider handles operational technology
Require coordination with engineering and equipment suppliers before changes that could affect production. Ask how assets will be discovered, how maintenance windows are approved, and how the provider avoids disrupting sensitive equipment.
NIST’s work on security segmentation in small manufacturing environments offers a useful starting point for separating systems by risk and function. Implementation still requires an understanding of permitted communications and operational constraints.
Review remote access and response coverage
Ask who can connect remotely, how access is approved, and how it is removed. Review authentication, session logging, and emergency support arrangements. Shared credentials and undocumented vendor connections make accountability harder.
Clarify what “24/7 monitoring” means in the contract. Does someone investigate the alert, contact a named plant decision-maker, and take an agreed action? Or does the service simply send a notification? Establish who has authority to isolate a system when doing so could stop production.
Test recovery instead of counting backups
List the data and configurations needed to restore critical operations. Include machine programs, application settings, licenses, and documented procedures where relevant. Assign recovery objectives based on business needs rather than accepting a default service tier.
Ask for a controlled recovery exercise with agreed scope and safety precautions. Record what was restored, how long it took, what dependencies were missing, and who approves corrective actions. A successful backup job does not prove the entire process can resume.
Compare proposals on operating responsibilities
- Named responsibility for office IT and production systems.
- Incident-response availability and additional charges.
- Log retention, reporting, and evidence access.
- Coordination with existing IT staff and equipment vendors.
- Contract exit support and removal of provider access.
Use a scenario such as unavailable scheduling software or a compromised vendor account during the evaluation. The provider’s response should make the first hour understandable. That is more useful than a promise that a collection of products will eliminate cyber risk.