When choosing a cloud provider, it pays to think small!

When you buy wine, do you go to the big discount store or the local specialty retailer? Chances are you do both, depending on the situation. The big-box store has selection and low prices, but the people who run wine store on the corner can delight you with recommendations you couldn’t find anywhere else.

The same dynamics apply to choosing a cloud service provider. When you think of cloud vendors, there are probably four or five company names that immediately come to mind. But if you Google rankings of cloud vendors according to customer satisfaction or relevance to small businesses, you’ll find quite a different list. There are hundreds of small, regional and specialty infrastructure-as-a-service providers out there. In many cases, they offer value that the giants can’t match. Here are five reasons to consider them.

Customer service – this is probably the number one reason to go with a smaller hosting provider. If you have a problem, you can usually get a person on the phone. Over time, the service provider gets to know you and can offer advice or exclusive discounts. The company just can’t match this personalized service.

Specialty knowledge – You can find apps for just about anything in the marketplace sections of the big cloud companies, but after that you’re pretty much on your own. If struggling with configuration files and troubleshooting Apache error messages isn’t your cup of tea, then look for a service provider that specializes in the task you’re trying to accomplish. Not only do you usually get personal service, but the people are experts in the solutions they support. They’ll get answers fast.

A smile and a handshake – There are several good reasons to choose a vendor in your geographic area. For one thing, government-mandated data protection laws may require it. Local providers also offer a personal touch that call centers can’t match. You can visit their facilities, meet with them to plan for your service needs and get recommendations for local developers or contractors you might need. Many small vendors also offer colocation options and on-site backup and disaster recovery. The technology world where sometimes everything seems to have gone virtual, it’s nice to put a name with a face.

Low cost – This sounds counterintuitive, but the reality is that many specialty providers are cheaper than the cloud giants. That’s particularly true if they specialize in an application like WordPress or Drupal, or in a service like backup. These companies can leverage economies of scale to offer competitive prices, then you get all the other benefits of their specialized knowledge. Shop around; you might be surprised.

Performance – If the primary users of the cloud service are people in your company and/or in your geographic region, you will probably realize better performance with a local vendor. That’s simply the law of physics. The farther electrons have to travel, the longer it takes them to reach their destination. This is particularly important if you plan to use services like cloud storage or if you need to transfer large files, an error-prone process that only gets worse with distance.

The IT Catch-22

OK, so everyone’s taking about it. Our industry is undergoing major changes. It’s out there. It started with a first architecture of reference with mainframes and minicomputers designed to serve thousands of applications used by millions of users worldwide. It then evolved with the advent of the Internet into the “client-server” architecture, this one designed to run hundreds of thousands of applications used by hundreds of millions of users. And where are we now? It appears we are witnessing the birth of a third generation of architecture, one of which is described by the IDC as “the next generation compute platform that is accessed from mobile devices, utilizes Big Data, and is cloud based”. It is referred to as “the third platform”. It is destined to deliver millions of applications to billions of users.

3rd platformVirtualization seems to have been the spark that ignited this revolution. The underlying logic of this major shift is that virtualization allows to make abstraction of hardware, puts it all in a big usable pool of performance and assets that can be shared by different applications for different uses according to the needs of different business units within an organization. The promise of this is that companies can and have more with less. Therefore, IT budgets can be reduced!

These changes are huge. In this third platform IT is built, is run, is consumed and finally is governed differently. Everything is changed from the ground up. It would seem obvious that one would need to invest in careful planning of the transition from the second to the third platform. What pace can we go at? What can be moved out into public clouds? What investments are required on our own infrastructure? How will it impact our IT staff? What training and knowledge will they require? What about security and risks?

The catch is the following: the third platforms allows IT to do much more with less. Accordingly, IT budgets are reduced or at best, flattened. Moving into the third platform requires investments. Get it? Every week we help CIOs and IT managers raise this within their organization so that they can obtain the required investments they need to move into the third platform to reap the benefits of it.

Don’t fall for marketing blurb

While watching a pickup truck commercial on TV lately, I couldn’t help but ask myself “How can all pickups have the best fuel efficiency in their category?” In a funny way, I hear the same in our industry with “most IOPS or terabytes per $”. It seems everyone’s the best at it. In one case, a client got the most IOPS per dollar he could get and he ended up having to change his whole data centre infrastructure because the IOPS he got were not of the right type!


In the storage industry, it seems that IOPS is the equivalent buzzword to horsepower (HP) in the automotive industry. So you’re going to try to get the most horsepower per $ when you purchase a car. You can get 350HP out of a small sports car or a pickup truck. Just don’t try to race with a pickup truck or tow something with the sports car! There’s a reason why you won’t see a Ferrari with a hitch! Though they both have 350HP, one has torque, the other one doesn’t. One is built for performance and speed, the other for heavy workloads. The same goes for data centres. Manufacturers will give you the IOPS you asked for and they can usually prove it! But do you know what IOPS type you’re looking for (sequential, random, read or write)? Why are you requiring those IOPS? Performance or heavy workloads? If you’re not sure, it’s an integrator’s core business and value to help you make sense of all the marketing blurb thrown at you, to help you choose wisely and protect your investment.

Charles Tremblay, ESI Account Manager