Is a multi-cloud strategy overkill for a mid-sized business, or is vendor lock-in a real threat?

All answers to this question.

For a mid-sized business, "true" multi-cloud—where you run the same app on two clouds—is almost always overkill and leads to massive technical debt. In 2023, a better approach is "Multi-Cloud by Specialty." Use AWS for your core infrastructure and legacy apps, but use GCP for your specific BigQuery analytics or AI workloads. This way, you get the best of both worlds without trying to make two completely different ecosystems talk to each other in real-time. Don't fight the platform; use its strengths.

   Answered 2025-11-07 by Melissa Bryant

Melissa, if we go with that "Specialty" approach, doesn't the data transfer cost between AWS and GCP eventually become more expensive than just staying on one platform?

   Answered 2025-11-09 by Steven Cook

• You're right, Steven. Data "gravity" is a real thing. If your data is in S3, moving it to BigQuery every hour will get pricey. The trick is to only move the processed results or use "Zero-ETL" integrations which are becoming more common. If your data volume is massive, staying in one "walled garden" is usually the more fiscally responsible move for a mid-sized company.

     Commented 2025-11-10 by Melissa Bryant

Vendor lock-in is a bit of a myth if you use open-source standards like Kubernetes or Terraform. It makes moving much easier if you ever actually have to.

   Answered 2025-11-12 by Jason Scott

• Spot on, Jason. Focusing on "Cloud-Agnostic" tools like Terraform allows you to stay flexible without the massive headache of a full multi-cloud deployment.

     Commented 2025-11-13 by Karen Walker