Why cloud IT providers generate profits by locking you in, and what to do about it - Part 1 of 2

The comparison has been made before, but the Cloud IT sector today looks in many ways much like the automotive industry of the early 20th century. 

During the first decade of the 20th century, more than 485 auto companies entered the market. After that explosive 1st decade a massive wave of consolidation followed, eventually resulting in dominance of just a few companies.  This type of industry evolution - technological innovation at the outset, followed by a wave of commoditization, and then finally consolidation is common in markets with economies of scale.  There will be differences though. 

Here are two predictions for differences in industry evolution between Cloud IT and 20th century autos:

  • The Cloud IT sector will consolidate an order of magnitude faster (perhaps two) than the auto industry did
  • Future Cloud IT oligopolists will work to maximize customer retention by maximizing switching costs (creating lock-in)


Cloud IT providers are subject to the same business rules as businesses in any other sector.  One of those rules is that, if you want to make long term sustainable profits (or “economic” profits for you MBA types), you must acquire customers and retain them.  Technology driven value add (sustainable competitive advantage) can help acquire customers, but as more and more IT becomes commoditized ,  customer switching costs start to drive the pricing required for cloud providers to sustain profits.

While auto manufactures used economies of scale, styling and brand loyalty to generate economic profits, cloud IT providers have to deal with a steeper commoditization curve (in part driven by open source) and a lower threshold (but still large) for reaching economies of scale (minimum point of the long-run average cost curve).  What this means is that commoditization happens more quickly as providers reach scale more quickly.  In this environment, a loss leader plus high switching cost strategy makes a great deal of sense.

 If you think about it for a moment, the prior generation of IT leaders (i.e. HP MSFT, IBM, CISCO, DELL, EMC, ORACLE, SAP) used a number of techniques to lock in their customers, so why not the next generation of IT leaders? 


How cloud IT providers lock you in


One of the key needs of 21st century enterprises is to be able to service customers better in the digital realm.  To do that they need to make sense of and monetize their data, all their data.  That data lives in lots of different systems and has to be integrated to be useful. Cloud services typically provide APIs (application programming interfaces) to make this possible.


Lock-in method one:  Enter the API

In the age of devops where everything in IT is being automated, cloud services are no exception.  By using cloud service APIs, different data sets and applications can be connected in useful ways.  An example use case is to use customer purchase information and social media activity to create a better support experience, manage inventory, and optimize pricing in real time.

But the more you automate, the more it costs to switch providers.

Cloud IT providers’ APIs are not standard in general, and often they’re highly complex For example, VMware had at least 5 distinct APIs for it vCenter application alone at one point. Lack of standards and complexity create additional rework, add to training and recruiting costs, and add significant time and risk to any transition to another provider.  This is true for any cloud provider that has a product with an API.

 In other words, complex, non-standard APIs drive profits for cloud providers.


Lock-in method two: Data Gravity

Data Gravity, a term coined by Dave McCrory (@mccrory) the CTO at Basho Technologies is a term used to describe the tendency for applications to migrate toward to the data as it grows. The idea is that as data grows, the difficulty of moving it to applications that use it becomes more difficult for several reasons, among them network costs, difficulties in accurately sharding the data, and so on.

 Data is growing faster than bandwidth.

There are two power law effects (similar to Moore’s law) going on with respect to data. First, driven by trends like social media and the Internet of Things (IoT), enterprise data is growing at an astronomical rate.  Second, network bandwidth is demonstrating robust growth as well, but is still growing much more slowly than data growth.  The implication? 

As time goes on, it gets harder to move your data from one provider to another.  Amazon was likely thinking about this when they launched AWS Import/Export Snowball at Re:Invent last fall.


Like the proverbial drug dealer, the business model for many providers is to provide small scale services heavily discounted or free as part of the cost of customer acquisition. While cloud is easy to consume, as scale grows, cost grows not just in absolute terms but in percentage terms.


In my next post, I’ll delve into the best practices that help you avoid (or at least minimize) cloud lock-in.


If you found this post informative, please share it with your network and consider joining the discussion below.

For more insights, sign up for updates at www.contechadvisors.com or follow me on Twitter at @Tyler_J_J