At first glance, there may seem to be very little difference between the major hyperscalers, particularly for organizations making their first foray into the world of Public Cloud or even just Cloud. But there are differences and they each have inherent strengths, so you need to identify the one best matched to your business, technical, and operational needs.
Based on my years of experience helping organizations across the public and private sectors migrate their SAP estates to the Cloud, I have developed a checklist to help enterprises choose the right Cloud provider for their business.
1. Culture
I’m a big believer that you’ve got to align and buy into the culture of the hyperscaler you choose because this is a long-term journey. Above all else, this is perhaps the one thing I would emphasize most to any potential buyer. Contrary to how it may seem when setting the business case, this is not a tactical choice – it’s a long-term strategic one. It’s actually pretty easy to move storage/compute from one Cloud to another, but once you get into the Cloud, you are going to be using many more capabilities than before and once those are embedded in your systems, it’s far harder to move away from them. As a result, you have to be aware that this is a long-term decision, so you must feel happy with the culture you’ve aligned with.
Some people naturally like Microsoft and are familiar with it, others are really impressed with Amazon’s retail focus and some people want the Google experience. I genuinely believe that culture is the most important factor – albeit far harder to quantify.
2. Cost
This is not just about the list price. Frankly, Amazon, Microsoft and Google keep each other honest when it comes to initial cost, which is far less of a differentiator, even though it will always be part of any business case. The second part of pricing is discounts – and understanding not just the initial discounts for migrating, but the long-term discounts available based on your volumes.
For example, Microsoft have made some moves in recent years to help organizations move their Windows landscape to the Cloud with a much better deal, offering far greater support than the others. In some cases, this can be the key decision factor. The other hyperscalers may have equally compelling offerings, but they cannot compete because of Azure’s obvious link to Windows. This is a good example of a very tangible thing to consider in cost modeling.
While they keep each other honest in terms of competitive pressure, some people are skeptical that prices will stay low, but I think we are seeing a fundamental move away from the ‘bait-and-switch’ approach to pricing previously seen in IT. This approach is aligned to culture as well so be sure to check what these IT giants have done in the past (and what are they likely to do in the future) that may affect your pricing extrapolations.
3. Feature set and innovation
If you’re a first-time user, it can be quite difficult to see any differences on features at all. Most of the core workloads look quite similar: storage, compute, networking, security etc. are generally the same, within a margin of error. But, once you get into things like data, Google are the colossus. Their data capability is impressive. Both Amazon and Azure also have very good data capabilities– better than you’ll find anywhere other than Google.
When it comes to Windows capabilities, as we have said Azure is way ahead in terms of license mobility, though both Google and Amazon can run Windows.
Where AWS are really differentiated is in machine learning (ML) and innovation, which has been breath-taking. The speed at which they release new things to the market is remarkable and many see AWS as being more customer-focused, due to their genesis and the position they take on innovation.
On first pass then, they may look the same but dig a little deeper and there are differences, and you should be sure to select the one that is most aligned to your own organization.
4. Unplanned downtime
The availability on any of the hyperscalers is without doubt much better than what can be achieved by any but of the most expensive on-premise alternatives. However, you may see some variations between the providers in different regions. These of course are hard to predict and all providers continue to improve this. However, you need to ensure you include architecture to mitigate against unplanned downtime in your target architecture and cost models.
5. Partner capabilities
It is worth looking at the capabilities of the hyperscaler’s ecosystem of partners. There is no manual for this, so you must figure it out either by yourself or with someone that has done it already. You need to look for partners with the right experience – and see if you can find those partners on each of the Cloud offerings. Some view this as secondary, but there are some organizations that follow the partner and go with their recommendation, rather than finding a partner to fit their choice of hyperscaler.
Choosing a hyperscaler platform should be approached with the same caution and diligence as any long-term relationship. The rules above can provide a very high-level framework to start with but then you need to dig deeper to formulate a detailed analytical assessment to determine which Cloud service provider(s) you will ultimately trust with your systems and data. You should also do a thorough analysis of your organization’s needs and expectations prior to comparing the hyperscaler candidates. This exercise is key to determining which hyperscaler is right for you and can deliver the features and functionality best aligned with your business and culture as well as your operational, security, and compliance goals.