Choosing between AWS, Azure and Google Cloud for an Indian SME

The three major clouds are closer in capability than ever, so the choice comes down to your team's skills, your data residency needs and where your existing licences already sit. Below is the framework we use in discovery.

1. Start with the workloads, not the vendor

List what will actually run: web apps, databases, batch jobs, analytics, ML. Each provider has strengths. Microsoft-heavy shops save on licensing with Azure; data-intensive teams often prefer BigQuery; AWS has the broadest marketplace and talent pool in India.

Mumbai, Hyderabad and Pune regions mean latency is no longer a differentiator. Compliance and cost modelling are.

2. Model three years of cost, not the first invoice

Free credits and startup programmes distort year-one numbers. Build a simple model with reserved or committed-use pricing, egress, support plans and the engineers you will need to operate it. The cheapest list price is rarely the cheapest total.

3. Check the compliance boxes early

• RBI and SEBI-regulated workloads: confirm data residency in Indian regions and the audit reports your auditors accept.
• Healthcare: look for HIPAA-eligible services if you serve international patients.
• Government and PSU work: MeitY empanelment narrows the field quickly.

4. Decide who will run it

A cloud is only as reliable as the team operating it. If you do not have in-house DevOps, choose the provider your partner knows best and insist on infrastructure as code so you can switch partners without switching clouds.