AWS, Azure, and Google Cloud are the three dominant public cloud providers. This guide explains what each does, how they differ on pricing, security, and integration, and what to consider when connecting them to the rest of your estate.
When planning a cloud deployment or migration, choosing between the major cloud providers is one of the first decisions to make. AWS, Microsoft Azure, and Google Cloud Platform are the three dominant options, each with genuine strengths and meaningful differences. Understanding those differences helps inform the choice, and understanding what they share in common is equally useful.
Amazon Web Services (AWS) holds the largest share of the public cloud market. It offers a comprehensive range of services across compute, storage, databases, networking, analytics, machine learning, and developer tooling. AWS has the broadest service catalogue of the three major providers, and its global infrastructure spans more regions and availability zones than either Azure or Google Cloud.
AWS suits organisations that want maximum service breadth and the largest ecosystem of third-party integrations. It has a strong community, extensive documentation, and deep specialisation in cloud-native workloads. The trade-off is complexity: the breadth of choice can make AWS harder to navigate without specialist knowledge.
Microsoft Azure is the second largest cloud provider by market share and the most widely adopted in enterprise environments where Microsoft products are already central. Azure’s strongest differentiator is its integration with Microsoft’s existing suite: Active Directory (now Microsoft Entra ID), Microsoft 365, Teams, Windows Server, and on-premise environments running Microsoft technology.
For organisations running hybrid estates, Azure’s ability to extend on-premise Active Directory into the cloud and provide consistent identity management across both environments is a practical advantage. Azure tends to be the default choice for organisations with significant existing Microsoft investment.
Google Cloud Platform (GCP) is the third major provider, with a smaller overall market share than AWS or Azure but particular strength in specific capability areas. GCP is widely regarded as a leader in data analytics, big data processing, and AI and machine learning workloads, drawing on Google’s own engineering heritage in these areas.
GCP is also increasingly relevant in healthcare, where its data platform capabilities and specific NHS-aligned offerings have made it a consideration for integrated care systems and health data programmes. For organisations whose primary requirement is analytics at scale or AI model training and inference, GCP merits serious evaluation alongside the other two.
All three providers operate on a consumption model: you pay for what you use, with options to commit to reserved capacity at lower rates. The details differ in ways that can matter at scale.
AWS charges for compute by the hour and offers significant discounts for higher committed usage. Azure charges by the minute and offers flexible short-term commitments alongside longer reserved instance options. Google Cloud offers per-second billing and has historically been competitive on sustained use discounts that apply automatically without requiring advance commitment.
Egress costs – what you pay to move data out of a cloud environment – are a significant variable across all three. For workloads generating high data volumes, egress costs can meaningfully affect the total cost of running in cloud. Private connectivity to cloud (AWS Direct Connect, Azure ExpressRoute, Google Cloud Interconnect) typically carries lower egress rates than internet-based connectivity, which is worth factoring into cost modelling for data-intensive workloads.
The most reliable approach is to use each provider’s own pricing calculator with your specific workload assumptions, rather than relying on general comparisons.
AWS IAM (Identity and Access Management) offers very granular control over user permissions and integrates tightly with other AWS services. The level of control is a strength for organisations that need precise permission management, but the complexity of configuring it correctly is a genuine learning curve. AWS-specialist skills are typically needed to implement it well.
Microsoft Entra ID (formerly Azure Active Directory) is widely regarded as simpler to deploy at enterprise scale. Its integration with on-premise Active Directory and its compatibility with third-party services and non-Microsoft platforms makes it a practical choice for organisations running hybrid or multicloud environments. For organisations already running Microsoft identity infrastructure, Entra ID extends naturally into the cloud.
Google Cloud’s Identity and Access Management follows a similar principle of granular control, with strong integration into Google Workspace for organisations using those tools, and a well-regarded approach to service account management for automated workloads.
AWS is the broadest platform and tends to be strongest for cloud-native development, serverless architectures, and organisations that want the widest possible range of managed services under one provider.
Azure is typically the strongest fit for organisations deeply invested in Microsoft tooling, running hybrid Windows environments, or where integration with Microsoft 365 and on-premise Active Directory is a priority.
Google Cloud is most competitive for data analytics, machine learning, and AI workloads, and for organisations where Kubernetes and containerised workloads are central to the architecture. Its BigQuery data warehouse and Vertex AI platform have genuine technical advantages in their respective domains.
In practice, many organisations use more than one. AWS for some workloads, Azure where Microsoft integration matters, GCP where data or AI capability is the driver. Managing that across a common connectivity and security layer is where the operating model question becomes important.
Choosing a cloud provider determines where your workloads run. How those workloads connect back to the rest of your estate – on-premise systems, other cloud environments, regulated networks, and users – is a separate decision that deserves the same level of deliberate design.
Each provider offers its own private connectivity service: AWS Direct Connect, Azure ExpressRoute, and Google Cloud Interconnect. These provide dedicated paths between cloud environments and on-premise infrastructure, with more predictable performance and lower egress rates than internet-based connectivity. For regulated organisations, they also make it considerably easier to evidence what is happening at the cloud boundary.
For UK organisations connecting cloud environments to HSCN or PSN, the connectivity architecture needs to be designed alongside the cloud choice, not after it. A healthcare technology company delivering a cloud-hosted product to NHS customers needs a compliant path from the cloud environment to the network their customers are on. That path needs to work for whichever cloud provider hosts the product.
Cloud Gateway delivers managed private connectivity across AWS, Azure, and Google Cloud as part of the Business Connect: Cloud product family, alongside compliant connectivity to HSCN and PSN for regulated organisations. All three can be connected under one operating model, with one contract and one team accountable for the paths between them. For more on how this works, see our Connect to the cloud page and Business Connect: Cloud.
