A client called me last year with a question that made me pause before answering: “Should we be on Azure or AWS?” I asked what they were actually trying to do. Turned out the answer was “run QuickBooks and host a WordPress site.” Neither of those needed a cloud platform decision at all: they needed a $12-a-month web host and an accounting subscription they already had.
That’s usually where this question goes wrong. Most small businesses asking “Azure or AWS or Google Cloud” haven’t actually confirmed they need to be choosing between raw cloud infrastructure providers in the first place. If you read my piece on what cloud computing actually is, you’ll remember that SaaS, the software you just use, like Microsoft 365 or Google Workspace, is 95% of most people’s real cloud experience. This article is for the other 5%: the point where your small business is actually running its own applications (custom, legacy, or maybe industry-specific), storing its own data at scale, or building something custom enough that “which cloud” becomes a real question instead of a hypothetical one.
When this question actually matters
You’re in real “which cloud” territory when at least one of these is true: you’re running a custom application your team built or commissioned, you’re storing and processing meaningful amounts of business data outside of an off-the-shelf SaaS tool, or you’ve got a development team that needs somewhere to build, test, and deploy. If none of that describes you, close this tab and go double-check your Microsoft 365 or Google Workspace backup situation instead, that’s the more useful thirty minutes.
If it does describe you, here’s the honest version of the comparison, without the vendor-brochure spin.
Azure: the pull is Microsoft, not the cloud itself
Microsoft Azure’s pitch to a small business is rarely “our cloud is better.” It’s “you’re already living in our stuff.” If your business runs on Microsoft 365, your identity and access already lives in Entra ID (what most of us still call Active Directory), and your team already knows Outlook, Teams, and SharePoint cold, Azure is the path of least resistance. Standing up a line-of-business application, connecting it to the same identity system your email already uses, and having your existing Microsoft licensing conversation just… extend, is a genuinely lower-friction experience than bolting a different vendor’s cloud onto a Microsoft-shop business.
That’s a real advantage, not a marketing one. Identity is the piece that trips up the most disaster recovery plans I see, I’ve written about that directly, and staying inside one identity ecosystem instead of stitching two together is one less seam that can fail you later.
Where Azure loses points: if you’re not already a Microsoft shop, there’s no compelling reason to start there instead of AWS or Google Cloud. You’d be adopting Microsoft’s ecosystem for its own sake, not because the underlying cloud platform demanded it.
AWS: the broadest catalog, and the steepest learning curve
Amazon Web Services is still the biggest player by market share, and it shows in the sheer size of the service catalog: if there’s a cloud service that exists, AWS probably has a version of it, often three versions of it. That breadth is a real advantage if you’re building something custom: whatever piece you need, it’s there.
It’s also the source of AWS’s biggest weakness for a small business: nobody accidentally becomes an AWS expert. The console is dense, the naming conventions are its own dialect, and it’s entirely possible to misconfigure something in a way that quietly runs up a bill nobody notices until it’s a five-figure surprise. I’ve told this story before, so I won’t retell it in full here, but an old AWS sandbox account of mine sat forgotten for a year or so and ended up mining crypto for someone else on my dime: real money, real lesson, and it’s exactly the kind of thing AWS’s flexibility makes possible if nobody’s watching closely.
For a small business, AWS makes the most sense when you’ve got technical staff, in-house or a trusted managed partner, who can actually drive it, or when you’re building something specific enough that AWS’s breadth of services is solving a real problem rather than adding real complexity.
Google Cloud: cheapest to start, thinnest safety net
Google Cloud tends to win the “cheapest to get started” conversation. Its always-free tier is the most genuinely usable of the three: a small always-on virtual machine, generous limits on things like Cloud Run and BigQuery, and per-second billing that doesn’t punish you for small or bursty workloads the way older billing models do. If your business leans on data analytics or anything in the machine-learning space, Google Cloud’s tools in that area are genuinely strong, arguably the strongest of the three.
The trade-off is support and market presence. Google Cloud has a smaller share of the enterprise market, which means a thinner bench of consultants, agencies, and hirable expertise built specifically around it compared to AWS or Azure. If something breaks and you’re on the free or lowest paid tier, you’re leaning on documentation and community forums, same as the other two, but there’s simply less of a Google Cloud-specific talent pool to pull from if you need to bring someone in fast.
For a small business, Google Cloud makes the most sense if cost efficiency and data/AI workloads are the priority, and you’re comfortable being a slightly smaller fish in a smaller ecosystem.
The part that catches people off guard: none of the free tiers are free forever
All three offer meaningful free tiers, and all three want you to think of that as “try before you buy.” It’s closer to a trial period with a trapdoor. AWS gives you 12 months of free access to a chunk of services; Azure does the same with a $200 credit for the first 30 days on top; Google Cloud pairs a small permanent free VM with a 90-day, $300-plus credit window. After that, you’re on pay-as-you-go, and pay-as-you-go bills don’t send a polite warning before they climb. They just climb.
I worked with a client once who accidentally selected the wrong SKU for a particular type of virtual security appliance. Instead of selecting the Bring Your Own License (BYOL) version, since they already owned licenses for those appliances, they selected the Azure Licensed version, where Microsoft bundles in its own license: two appliances at $5,000 a month each instead of $350 a month each on the BYOL version. Little things like that can bite you very unexpectedly, and it’s only a wrong button click away. (I’ve got an article coming up on cloud costs and cost optimization that gets into this much more in depth.)
Budget for what happens after the free tier before you start, not after your first invoice posts. And if you’re a genuine startup rather than an established small business, all three platforms run credit programs worth looking into before you assume you’re paying full freight from day one: AWS Activate, Microsoft’s Founders Hub, and Google for Startups all exist specifically to subsidize early-stage companies, sometimes substantially.
Support isn’t free either
This is the one small businesses forget to budget for entirely. The free support tier on any of these three platforms gets you documentation and community forums: nothing else. No phone line, no guaranteed response time, nobody to call at 9pm when something’s down. Actual technical support starts around $29 a month for basic email support with next-day response times, and climbs fast from there if you want 24/7 phone support with an hour-or-less response window. If your business genuinely can’t afford downtime, that support tier isn’t optional: it’s part of the actual cost of the platform, and it should be in the budget conversation from the start, not discovered during an outage. And again, not to be salesy here, but in many cases if you don’t have a technical staff that is at least somewhat cloud savvy, having a good MSP (Managed Services Provider) that you have partnered with builds in a safety net far more reliable than hoping a forum post has the right answer.
So what do you actually do with this
If you’re already a Microsoft 365 shop and you’re standing up your first application or data workload, start with Azure. The identity and licensing continuity is worth more in practice than most feature-by-feature comparisons suggest.
If you’re building something custom, have technical staff who can drive it, or expect to need a specific service that only one provider offers well, AWS’s breadth makes it the safer long-term bet, just budget real time to actually learn it, or budget real money to hire someone who already has. A “we’ll figure it out as we go” mentality here will cost you way more than being proactive about it.
If cost efficiency and data or AI workloads are the priority, and you’re comfortable with a smaller surrounding ecosystem, Google Cloud is worth serious consideration, especially for a leaner team that wants to move fast without a steep learning curve.
And if none of the above describes your business yet, if you’re still running SaaS tools and wondering whether you need “a cloud,” you probably don’t. Not yet. Revisit this once you’re actually building or storing something that needs it.
The takeaway
The Azure-versus-AWS-versus-Google-Cloud question is a real one, but it’s usually asked several steps before it needs to be. Confirm you’re actually in the 5% first. If you are, the right answer has less to do with which platform is “best” and everything to do with what you’re already running, who’s going to operate it, and what you’re willing to pay when the free tier runs out.