When the cloud stops: your supplier’s outage becomes your loss

The decision behind the headline

This review focuses on hosted applications, cloud outages, service levels and customer obligations.

Trace the online retailer’s checkout, payment gateway, identity service and order database. If several rely on the same cloud region or provider, a second storefront may not provide meaningful independence. Ask which failure stops transactions and which merely delays reporting. A merchant may still receive orders through a fallback while fulfilment remains blocked, producing a different loss from a complete outage. Test a manual process with actual staff and a small sample of orders rather than assuming a spreadsheet is sufficient.

A hypothetical scenario to test

An online retailer cannot take orders because its external platform is unavailable while local systems work. This is an invented example, not a reported claim or a promised payout. Reconstruct the sequence from the first sign of trouble to ordinary service resuming. Ask what would happen if the problem occurred during a holiday or at the busiest time of year. The stress test should expose assumptions about availability and timing before they are relied upon.

Request supplier interruption wording that names the applicable services and event triggers. A malicious attack, a technical outage and a scheduled maintenance window should not be treated as interchangeable without confirmation. Ask when the waiting period begins and which documentation of provider downtime is required. Preserve provider status notices and your own order timestamps. Compare sales recovered after reopening with orders abandoned permanently; do not turn every delayed transaction into a lost sale. The insurance discussion should also explain whether service credits reduce the claimed amount and whether the fallback’s additional expenses are separately eligible. A supplier promise of uptime does not determine your insurer’s response.

Identify the contractual boundary

Service credits promised by a cloud provider can be much smaller than your operational loss. Read them as contractual remedies, not substitutes for business-interruption cover. Test whether the supposedly independent fallback uses the same identity service or internet connection as the failed application.

Compare the wording, then the price

Use an identical scenario with each insurer and request clarification of dependent interruption, named suppliers, non-malicious outages and time thresholds. Ask for the relevant clause and a written explanation of anything left uncertain. An optional extension should appear in the schedule, not merely in a sales discussion.

Reduce the exposure before transfer

A useful first action is to export essential data and test a manual order process independent of the main provider. Assign an owner, a completion date and a simple test of whether the measure works. Written procedures matter only if staff can find and use them under pressure.

Evidence and the next review

Prepare supplier contracts, outage notices, order history and a map of shared dependencies. Date the records and keep an accessible copy away from the affected process. Keep offline access to the insurer contact, incident plan and supplier escalation list. A claim notification does not replace any required regulatory notification. Review the arrangement when your activity, technology, locations or counterparties change.