Multi-cloud
Also known as: multicloud, multi-cloud strategy
Multi-cloud is a strategy of using cloud services from more than one provider, such as running some systems on Microsoft Azure, others on AWS, and analytics on Google Cloud.
Organizations pursue it to avoid dependence on a single vendor, use each provider's best services, meet regional requirements, or improve resilience. While multi-cloud sounds safer, it multiplies the skills, tools, and security controls needed, and many organizations end up multi-cloud by accident rather than design.
Multi-cloud, explained at three levels
At home
Multi-cloud is like shopping at several grocery stores, buying each item where it's best or cheapest, instead of relying on one store for everything.
At work
Using more than one cloud provider can reduce dependence on a single vendor, but adds management effort.
At home
Using Google Photos, iCloud, and OneDrive at the same time is a personal form of multi-cloud, with some files in each.
At work
Multi-cloud requires consistent identity, monitoring, cost management, and security controls across providers. Infrastructure as code tools help standardize deployments.
At home
Spreading files across too many services makes them hard to find and back up. In my opinion, choose one main service and one backup.
At work
Decide where portability truly matters, avoid lowest-common-denominator designs that waste each provider's strengths, budget for data transfer between clouds, and maintain skills for each platform you use.
Real-world examples
At home
Aisha stores work files in OneDrive for her job and personal photos in Google Photos, a simple version of using multiple clouds for different needs.
At work
Marcus is IT manager at a Winnipeg insurance company. They run core systems on Azure but use Google Cloud for one analytics tool its data team preferred.
Common confusions
- Multi-cloud automatically protects against outages. — Only if systems are designed to fail over between clouds, which is complex and costly.
- Multi-cloud always saves money. — Extra tools, skills, and data transfer fees often increase total costs.
For analysts
When multi-cloud is proposed, I'd ask for the specific risk or benefit it addresses. Vague 'avoid lock-in' goals rarely justify doubled complexity.
History & origin
As cloud providers matured in the 2010s, organizations increasingly found themselves using several, often because different teams chose different services or companies merged. Tools such as Terraform, released by HashiCorp in 2014, and Kubernetes made it easier to manage resources across clouds. Concerns about lock-in and high data transfer fees fuelled debates, and regulators in the UK and Europe examined cloud competition from 2023 onward. Major outages, such as the large AWS disruption in October 2025, renewed interest in spreading critical systems across providers. In my view, multi-cloud works best as a deliberate choice for specific reasons, not as a blanket insurance policy.