From the #Cloud: #Azure vs. #AWS – my twopence

By | December 31, 2014

I’ve been doing some consulting for a couple of global organisations recently, with the emphasis on “who’s da boss?” when it comes to Cloud platforms.

A bit of GoogleBing-ing will get you lots of opinion on this; some from the vendors themselves, some from employees, some from folks that clearly have a vested interest in one or the other of the platforms, some from consultants and some from “the press”.

All-in-all, it’s pretty biased one way or the other. Trying to get reliable, independent opinion – based in fact – about which platform is “better” is almost pointless. It is almost non-existent. This frustrates me hugely.

The possible exception is the Gartner viewpoint. A research organisation, Gartner is independent (although many would argue against this and there is evidence to suggest less than perfect behaviour in the past) and relied upon by CIOs around the world (although again, it worth noting that it has been historically suggested that Gartner “influences the IT industry more than it observes” which implies its bias can be altered if commercially appropriate which has resulted in accusations of “endorsement-like output” as opposed to “research output”) when they are making both strategic and tactical technology investment decisions.

The Magic Quadrant viewpoint on IaaS essentially ranks only two players worth speaking of – AWS and Azure; ranking AWS way above Azure in terms of being “better” (i.e. more “top right”). It’s tough to argue against this in absolute terms; the methods Gartner use are scenario based and reasonably fair, but it is worth noting that in the same way a vehicle service is not an indication of roadworthiness, research papers such as Magic Quadrant reports (and this post for that matter) are a point-in-time view and tend to be out of date the minute they are published.

If you follow the trend (of Azure vs AWS) in Magic Quadrant over the last few years, AWS has had its strong leadership position for several years now and the only challenger (currently) closing in is Azure – and at an alarming (from Amazon’s perspective) rate – but nonetheless at this time and by this measure, AWS is “the best”.

Anyhoo, back to my frustration.

One of the principles of being a consultant is in being able to provide unbiased (unless bias is appropriate) advice and opinion so that customers can make the best choices. Hard to do given that (a) it is almost impossible in today’s fast-moving technology landscape to be a master of all things Cloud and (b) when we turn to our wider vendor, community and networked peers the opinion is largely one-sided.

The reason it frustrates me so is that it is actually quite simple to be unbiased. All that is needed is more focus on customer needs. Each platform (specifically in my recent context of Azure ‘vs’ AWS (and yes, I do recognise and accept that other platforms exist but I have been focused in this comparison space due to existing customer investment in both platforms)) meets different client needs in different ways and so is “better” in the eyes of different customers based on their needs.

If we ignore the arms race of the “features” of Azure and AWS (kinda how Gartner works) and focus on customer scenarios, we can draw conclusions about which platform would be better for a given need and step away from arbitrary “XXX is the best” type thinking.

Let me provide a trio of examples.

Example 1

Hip, funky “Web 3.0” startup with the next “new” thing for the web. Let’s call them “iFaceTwittWhatsaGram”. Self-funded until round 1, they have aggressive growth plans based around a lightweight, high transaction app idea for the mobile web, you get the picture.

Example 2

An established, multi-national corporation with 30,000 employees in 10 countries. Named “GlobalMegaGrower”. They’re heavily invested in existing on-premises IT, and rely heavily on home-grown applications built in .NET.  It would not surprise you that they are security conscious but not overly so. Corporate to the core.

Example 3

A mid-tier international financial services organisation with 2,500 employees. “CashMoneyTrading Inc.” is a growing enterprise that services financial markets in Europe and the USA. Highly regulated, they’re deeply mindful of security and sovereignty. They’re entrepreneurial but have an enterprise mindset as they’re on the glide path to IPO.

For iFaceTwittWhatsaGram, AWS is likely to be far and away a better option. Scalable, reliable, inexpensive to enter and grow and DNA ingrained with the open source mindset of Linux-land. It’s a match made in heaven.

In the case of GlobalMegaGrower, Azure will all day be the better option. Tightly-coupled hybrid capabilities, seamless on-prem to Cloud integration and migration added to native, expert support and understanding of the enterprise applications in use across a business that is of a size that will attract discount levels to even up the pricing disparity. It’s a total no-brainer.

When it comes to CashMoneyTrading Inc. the points of differentiation become less obvious and our ability to state that one platform or the other is better is diminishing. The hybrid nature of Azure is offset by the cost benefits of AWS. Both platforms have potential challenges with their security/sovereignty that could be cause for concern despite both platforms being inherently secure enough for most organisations.  The field is level.

Tricky, no?

Although I am being brief (clearly, in depth exploration and analysis with CashMoneyTrading Inc. might uncover something specific that would lean them towards one or other of the two platforms) we can see that in this situation it would be incredibly arbitrary and somewhat daft to accept that “XYZ is better”.

My point is that ignoring the arms race of features and benefits is key when considering the Cloud.  Match requirements to what can be delivered and that’s what’s “best”.  It’s simple.

Now for the biased bit… 😉

If put on my Microsoft bias hat, this is where the pieces of the pie outside of Azure PaaS/IaaS are swinging customers towards the Microsoft powered services and solutions.

For full disclosure, in the projects I mentioned at the top of this post, 100% of the customers chose the Microsoft route to progress.  Not due to excessive Jedi mind-tricks on my part, but because the pie when looked at from 100,000 feet simply looks tastier with Azure at the centre surrounded by all the Microsoft sugary, creamy and custardy accompaniments:

  • Office 365 for messaging, unified communication and “simple” collaboration
    • Enterprise IT “in-a-box” ready to go for 90% of use cases with Cloud scale and affordable pricing
  • Active Directory (with sync services) for authentication management
    • Authentication for service access regardless of the location of the user or application
  • Enterprise applications and application components
    • SQL Server, BizTalk, SharePoint and Exchange provide unparalleled functionality in on-premises, IaaS hosted or hybrid scenarios
    • Visual Studio and .NET ensure that organisationally specific requirements can be delivered in future-proof, packaged and supportable ways
  • Money, ca$h, moola. Micro$oft is a financial bear. $400B in mkt cap and in excess of $100B in cash.  Compare that to Amazon with $140B and less than $10B in cash.
    • Microsoft is enjoying epic profits despite huuuge investment in re-inventing itself as a devices and services business
    • Microsoft can simply out-invest Amazon. I would go as far to suggest that Amazon could literally go broke trying to keep pace with the Redmondian beast both in terms of capability/capacity (datacenter) and price per unit (sales)
  • Footprint
    • Microsoft is dominant as it is still benefiting from first-mover advantage it established in the 80’s
    • A ridiculous percentage of servers and desktops run Microsoft OS, it’s bloody everywhere
    • Microsoft Office rules the roost.  Even iPad users have it.  ‘Nuff said
    • Deploying, migrating to and skilling up in new tech is risky and expensive – why would a customer not in a green or brown field position even bother?

There are other delicious extras, but it would be churlish to embark on a “Microsoft Rules” post as that is not the intent here.

I guess in summary I am trying to say that just because the Cloud is “new” (sort of) the rules of how customers need to select technologies have not fundamentally changed.  Sure, the choices have been updated.  What’s possible is always evolving, and always will be.  The key is, has been and always will be for choices to be made based on best fit to requirements.

Simples.

more to follow…

One thought on “From the #Cloud: #Azure vs. #AWS – my twopence

  1. Peter Ridgway-Davies

    I have been an advocate of Microsoft for 30 years. They brought stability to the PC/networking technology space and have been a very safe bet. However this is changing. Microsoft are defining their own cloud marketplace using money from the on premise corporate cow. I see all those billions of dollars being spent on developing cloud products and far less on the on premise platform. It would be fair enough if this was driven by what customers are demanding but this is not the case. It is being driven by Microsoft’s desire to create a new niche for themselves. Most of my clients are corporates and they have huge issues chiefly security, performance (how on earth can a cloud solution match on premise?) and cost! It costs money to put stuff into the cloud. It is not free so when people decide to move to the cloud then there is the additional cost involved.
    Cloud benefits some businesses – those that are small and seeking to rapidly build a great solution when they have little invested in their current IT. It does not suit everyone and those who hesitate risk being stuck on an infrastructure that will be starved of investment. Hybrid offers a partial solution but how long lived this is I don’t know.
    For me the two things stopping me are
    1. Ownership of the data – currently the US legal system believes data stored on Microsoft owned hardware anywhere in the world is under their jurisdiction. With such a threat how can we store confidential data on the web?
    2. Latency – how on earth can a system based on the web match on premise speed and responsiveness?
    I could go on. Of course there are also many benefits of the cloud but for me these are outweighed by the negatives.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *