The benefits of cloud native Vs cloud wrapped software

Realizing the benefits of true cloud computing applications

Cloud computing is changing the way we all consume and interact with technology. Whether this is at home with Netflix, or with devices such as Amazon Alexa, or at work, with cloud-based services like Office365, Salesforce or other specialist enterprise applications. In business, cloud computing is an enabler; it creates strategic change in the way businesses of all sizes work with technology. In short a real cloud native service will make your life easier, more efficient and less expensive.

The benefits of real cloud computing are clear, but many providers simply wrap legacy applications in cloud based hosting. These cloud wrapped applications often come with many of the problems of legacy applications. They’re also harder for your vendor to support and maintain. Innovation is limited and higher costs will eventually be passed to you, the client. So, what are the benefits of cloud native applications and what are the problems with cloud wrapped solutions?

 

Cloud Native

Cloud Wrapped

Vendor benefits

Client benefits

Vendor problems

Client problems

Operations

  • One set of code for all clients
  • More reliable service
  • Multiple installations mean that branch versions proliferate
  • Unreliable and difficult to service

Upgrades

  • All clients upgraded simultaneously.
  • Reduces development costs
  • 100% of developers working on new version
  • Increased innovation
  • More reliable
  • Free upgrades (no hidden costs)
  • Always using the latest version
  • Each upgrade is a project for each client
  • More clients = more complexity and more risk of failure
  • Need to have permanent teams of developers supporting old versions
  • Upgrades are expensive with direct and internal costs
  • Risky
  • Infrequent
  • Supplier costs end up being charged to the client

Capacity

  • Elastic computing matches supply to demand
  • Never worry about upgrading hardware – it just works!
  • Have to frequently upgrade hardware
  • The client must agree first
  • Once agreed, the lead time to implement can take weeks
  • Service is often slow because there isn’t enough capacity during peak loads and calculations

Connectivity

  • Linking to any system with an API is straight forward
  • Developing one API for all client data extractions is much easier than supporting multiple file extractions
  • Easy programmatic access to output data
  • Plug directly into corporate data management strategy
  • No manual re-formatting of extracted data
  • Maintaining data extractions for each client is difficult
  • High support costs
  • Without a Web API, extracting data involves manual processes
  • Increased operational risk
  • Doesn’t link directly with other on-premise applications

Security

  • Designed from the ground up as part of the software architecture
  • Granular controls for each aspect of the service allow for sophisticated roles based permissions
  • Easier to externally audit
  • Means more users can access the system seeing only what you want them to see
  • Analytics can be shared across the business and even outside to clients
  • Easier to perform data security due diligence
  • Security features often added as an afterthought
  • Difficult to audit and prove to clients the security architecture
  • Hard to implement modern encryption techniques on legacy code
  • Not designed for the cloud so data security is a concern
  • Poor permissions structure means it’s harder to share the application across the business
  • Difficult to pass internal due diligence process

Some of the cloud native productivity gains in numbers

Upgrades

300 clients

Each upgrade costs on average $20k

Each upgrade requires 3 months on average to achieve sign off.

Approximately 15 clients can be upgraded per quarter with existing resource.

Therefore is takes 5 years to upgrade all 300 clients to one new version at a cost of $6.0m to clients.

With Revolution there are 5 upgrades a year, all free and instant.

In 5 years, we can effect 25 upgrades = 25 times more efficient and zero cost = infinitely more efficient.

 

Development costs

Our development team has 60 people. Over five years we may spend £15m on developer salary costs. How much of that money and time is focused on new functionality versus supporting old versions of software?

Before StatPro Revolution, only 60% of development time was spent on new features, 40% was spent on on supporting older versions.

So over five years, £6m is spend on support time and £9m is spent on developing new functionality.

With Revolution all developers work on new features with support taking only 5% of the time.

Therefore cost of support over 5 years = £0.75m.

Money spent developing new software and functionality over 5 years = £14.25m. This allows for a dramatic shift in software delivery and innovation.

 

IT Costs

A typical client would need three environments

  1. Production
  2. UAT
  3. Disaster recovery

Hardware and software licences have to be decided on and then implemented by the client or by the vendor.

Each installation ends up being unique and will cost between $50k and $300k per annum to run. Every 3 years, the hardware will need to be refreshed.

Hardware, software, engineers, project staff, procurement, business input, management, external suppliers, delivery, co-ordination are all necessary and create hidden costs.

Revolution programmatically summons additional servers from AWS as and when required without the need of IT engineers. When the servers are no longer required they are released. A 40-core server stack might cost $8 for one hour. To rent this for one year would be $71k, but as it is only needed for one hour per day it would be less than $3k per annum.

 

Read More

How to tell the difference between cloud native and cloud wrapped software