Last checked: 7 November 2026
In short: A hyperscaler is a very large cloud provider that runs huge data centres around the world and rents computing power, storage and AI services to other businesses. AWS, Microsoft Azure and Google Cloud are the main three.
Table of Contents
How it works
- Build at scale. The provider runs hundreds of data centres filled with servers and AI chips.
- Rent by use. Customers pay for the storage, computing time and AI models they use, often by the second or by the request.
- Add services. On top of raw capacity, it sells ready-made tools such as databases, security and AI models.
Where you’ll meet it
Most business software, from your accounts package to your CRM, runs on AWS, Microsoft Azure, Google Cloud or Oracle Cloud. Many AI tools you buy are built on one of them.
What it means for your business
- AI arrives through them. Hyperscalers offer models from several AI firms, so you can test options without new contracts.
- Check your exit costs. Moving data out of a cloud can carry fees, so ask about them before you commit.
- Use your bargaining power. Larger contracts can win discounts, credits and better terms, so compare at least two providers.
Watch out for
- Lock-in. Licences and fees can make it costly to switch providers or use two at once.
- Data location. Check which country stores your data, as this affects privacy rules and contracts.
Example
Competition and Markets Authority. On 31 July 2025 the UK competition regulator published the final decision of its cloud services market investigation. It found that competition was not working well and that AWS and Microsoft each held about 30% to 40% of UK cloud spending. It recommended that both be considered for “strategic market status” investigations. In March 2026 the CMA opened one into Microsoft’s business software, after both firms set out steps on switching fees.
Hyperscaler vs compute
| Hyperscaler | Compute | |
|---|---|---|
| What it is | A company that sells cloud services | The processing power itself |
| What you get | Computing, storage and software tools | Capacity to train and run models |
| How they relate | A major seller of compute | One of the things hyperscalers sell |
Frequently asked questions
Should a small business work with a hyperscaler directly?
Usually not at first. Most small firms reach a hyperscaler through software they already use. Going direct makes sense once you build your own apps or store large amounts of data.
How hard is it to switch between hyperscalers?
It depends on how much you use their own tools. Plain storage and servers move fairly easily, while built-in databases and AI services often need rework.
Are smaller cloud providers a safe choice?
They can be. Check their security certifications, where they store data and their financial health, and make sure you can export your data if they close.