The Great Cloud Vendor Lock-In Fallacy

Some believe Cloud vendors operate in a Wild West-like corner of the Internet where anyone with a domain name and an AWS account can set themselves up and attempt to lock you into their systems for years to come, by simply keeping their platform as proprietary as possible. This simply isn’t the case - Cloud vendors are in most cases quite reputable and treat your data as sacrosanct. There is already a lot of support for standards within the cloud market and many vendors are building standards-support into their environments without the need for regulation. Openness and transparency are inherently easier for a Cloud vendor to achieve given the availability of metadata within their environments. Try delivering openness and transparency in a network of closed systems and platforms.

I had the opportunity to write a post for the Future of Cloud Computing Forum; I'd like to share the post here with you.

Most innovations and disruptive technologies tend to bring out what I refer to as the flat-Earthers – individuals who wait for the tipping point of a new technology or idea to be well past proven before getting on board, or as Geoffrey Moore calls them in 'Crossing the Chasm' - "Laggards!" Why is this? It’s because as humans we’re very dependent on habit forming behavior and love to hark back to a “better time” – and ”they don’t make them like that anymore” thinking. Adapting to change takes time, and I believe that Cloud Computing is at last winning over some of the last and most ardent deniers. But there may still be one last fallacy to overcome, which is cloud vendor lock-in. Nothing is quite as open and flexible as an old on-prem solution – or is it?

When mainstream Cloud Computing appeared (setting aside the whole mainframe, client/server, cloud discussion for a moment), there was much worry about how secure and ‘safe’ this new Cloud environment would be. Security was the cloud-deniers main argument for staying firmly entrenched in their onsite infrastructures. Now most agree that the Cloud generally offers a greater degree of security and resilience than would be possible on site, unless you have a DoD sized budget. Unfortunately there are still those who would have you believe the Cloud means you’re locked into a cloud-vendor, until death you do part.

I believe that this Cloud vendor lock-in fallacy is being touted by those who have a vested interest and would rather you kept your data onsite and within your own data center. There are many reasons cited by these flat-Earthers. Let me try to dispel their most common arguments.

Cost of migrating off Cloud platforms One of the two most common problems cited is the cost of migrating away from Cloud platforms when the customer chooses to leave or the Cloud vendor implodes (implosion, another symptom of this Cloud computing sickness apparently). The argument usually goes like this; companies sit down to work out their ROI on a Cloud investment and get excited to see the Cloud demonstrably saves them money. However in their excitement the IT team neglects to factor in the cost of migrating OFF a cloud platform at some time in the future. This cost will come as a surprise when the inevitable Cloud Rapture finally arrives. Complex deployments of information-worker software always seem to need someone to migrate data during an upgrade or swap out. I would argue that the cost of migrating from one on-premise solution to another is likely to be dramatically more expensive than a Cloud solution, simply because the Cloud vendor is:

  • expecting to absorb or ingest your data at some point, and
  • already planned on giving your that data back at the termination of your contract.

No customer should sign a Cloud vendor’s contract without that clause, and no Cloud vendor should expect them to. On premise solutions hide all sorts of complexity that makes it very hard for you to leave that solution. The cost of finally moving to any new platform means you’ve created your own on-premises vendor lock in.

Cloud Standards or Lack Thereof

The second of the two loose threads regularly pulled in this discussion: Industry standardization within the Cloud market is still a ways off. But this is not for want of trying. To address this, organizations like Open Stack counter the problems caused by a lack of Cloud standardization and drive the concept in the right direction.

The Highly Customized Nature of On-Premises Applications

Customers using big CRM and ERP applications are a breeding ground for custom modules and plugins for those monolithic applications; it’s the only way they get the functionality they want. For the time being it’s also the primary way the CRM and ERP vendors hold onto their customers, but the groundswell behind Cloud platforms like Salesforce mean this won’t last forever.

This corner of the discussion is quite closely tied to the cost of migrating a platform to or from a Cloud vendor too. On-premise applications that have undergone extensive deployments across massive WAN enabled infrastructures, and then have had to be tweaked with endless customizations, are going to be almost impossible to migrate to any other platform, Cloud or not. Declaring system bankruptcy and starting from scratch is often their only way out.

An example that demonstrates this well, might be moving a DB2 system to Oracle over the weekend.

Proprietary Formats and Data Types

Proprietary system formats and storage data types are possibly the earliest form of smoke and mirrors used by on-premise software vendors to lock their customers into on-premise solutions; odd that many of those on-premise vendors have since launched or supported some sort of cloud-washed version of their own platform.

Luckily the truth slowly prevailed as openness and transparency became the chosen path for reputable Cloud vendors, who in a bid to make their platforms more attractive and capable than their on-premise competition built in the functions require to search, export, extract and generally make your data available when you want it.

The Final Fallacy?

Quite simply, I believe the last great fallacy touted about Cloud computing is that of cloud vendor lock-in. Cloud is innovative and is becoming quite disruptive - businesses are turning to the Cloud to help solve complex problems which only a few years ago would have required a huge on-premise infrastructure.

We won’t have to wait long for this fallacy to die away. As more and more enterprises consider the Cloud first over on-premise solutions, the ruthless due diligence of the end users, administrators and the rest of the IT community will filter out these last remaining pieces of flat earth rhetoric.

Photo CC via ||read|| on Flickr