Skip to main content

The evolution of cloud adoption in Big Data

In the last decade, cloud-adoption in the Big Data space has been rising exponentially. With cloud platforms gaining increasing popularity over their on-premise counterparts and the nature of data evolving from static resource to business lifeblood.

Common connectivity - a catalyst for the rise of cloud computing 

Businesses have been migrating toward a hybrid cloud infrastructure for a while now. Hybrid cloud strategies allow for a better balance between each organization’s mixed cloud infrastructure needs since they combine both private and public strategies.

The evolution of cloud computing began with the need to freely share information and the introduction of common connectivity standards and frameworks. This required interconnectivity across devices, creating a difficult task for on-premise data solutions, where every hardware manufacturer had its own software protocols.

The answer to interconnectivity… cloud computing, more efficient, productive, agile, and secure - the benefits of cloud-based computing far outweighed existing on-premise solutions.

Over the following decade, cloud computing gradually became the standard in the Big Data space. Enterprise migration to cloud-based big data platforms accelerated, with COVID-19’s demands for a distributed, remote workforce putting cloud computing solutions into overdrive. It became clear that enterprises could no longer afford to have any information location bound, siloed away, or managed on-premise. 

Data as a “flow of information” vs a static resource

Over the past few years, we also witnessed a shift from viewing data as a static resource to viewing it as an information flow that must be kept in motion.

At the same time, IT has evolved from a resource-intensive cost centre, to a core value driver and has risen to prominence as integral to enterprise value creation, asset management, and business protection.

These trends led to the acceleration of cloud-adoption across all industries, as the world began to appreciate the significance of data, with its value coming from its ability to flow, interact, and be exchanged.

That’s not to say that on-premise doesn’t still have a place in Big Data. Popular solutions for on-premise data storage include data lakes that operate on Hadoop. These Hadoop-managed data lakes offer an extent of scalability, the ability to house different types of data, and some analytics capabilities.

However, large enterprises are likely to have multiple data lakes outside of HDPS and other data initiatives, all of which must be processed. Which is why many have gone with a hybrid approach - utilising both on-premise and cloud solutions to store and process their data.

Cloud adoption then and now

Early adoption of cloud computing was primarily restricted by security concerns and poor trust. Even when the cost and efficiency benefits were clear, enterprises would sometimes choose to keep data on-premise and manage it themselves to ensure security.

Ironically, security is now a prime advantage of cloud-based big data platforms.

Cloud-based platforms’ adherence to regulatory standards serves as a major trust creator. They also now offer compliance as a service. This further relieves the enterprise's IT management burden and frees up resources to be used on core value creators. 

Cloud adoption today

In our recent Big Data Report, we surveyed we asked 100 Big Data strategists, architects and users across industries and asked them about their big data and cloud strategies.

The majority have opted for a hybrid approach, not quite ready to take the plunge into a cloud-first strategy but more than willing to take advantage of the many benefits the cloud has to offer. The second largest group were made up of cloud-first and multi-cloud adopters with only 5% of those surveyed remaining fully on-premise.

Cloud storage and computing have made huge strides in the past few years, largely thanks to the main cloud service providers, like Amazon Web Services (AWS), Microsoft Azure and Google Cloud Storage. These companies have been investing heavily to develop robust cloud infrastructures, tools and applications to offer as an affordable and flexible service to the world. It therefore comes as no surprise that the majority of our respondents have incorporated the cloud in their data strategy, either with a cloud native, multi-cloud or hybrid approach.

Cloud adoption

For 2021 and beyond we expect an increasing number of companies embracing a cloud-first strategy as the amount of data increases, and costs associated with space, power usage, infrastructure and security become overwhelming. Most cloud solutions are highly scalable and help to reduce infrastructure cost and complexity. 

Cloud migration risks

There are still some risks associated with moving enterprise infrastructure to the cloud.

  • Security still needs to be carefully tended to. Security protocols must be based on cloud-native best practices, not duplicates of on-premise management habits.
  • When project development and deployment is democratised across the organisation, user access management becomes a core area. Users who make unintentional accidents can be as destructive as malicious saboteurs. Cloud migrators need to ensure that the human capital factor is well managed.
  • Enterprises also need to ensure the right skill, when moving their infrastructure. Gartner estimates that through 2022, insufficient cloud IaaS skills will delay half of enterprise cloud migrations by two years or more. A cost optimised enterprise migration depends on sourcing skilled cloud professionals to successfully meet adoption objectives.

3 low-risk approaches to cloud migration

Managing distributed, democratized computing resources at scale, can be optimised by creating a centralised control centre or team. Cloud implementation, maintenance, and general operations can then be controlled from this hub.

There are three risk-reducing approaches for a secure enterprise-level cloud migration. These follow the dev-ops method and are suitable for different timelines, resource levels, and environment complexities:

  • Life and Shift. This is simpler, with less ambiguity and risk. It’s better when there are tight deadlines, such as if the data centre lease is expiring. However it’s not always the most cost effective and existing architecture doesn’t always map into a cloud solution.

  • Re-Architecting (lift, tinker, and shift). This is ideal for maximizing the benefits of cloud migration and bringing the most ROI. It requires research, planning, experimentation, education, implementation, and deployment. This is time and resource intensive, but pays off with reduced hardware, lower storage and maintenance costs, along with increased flexibility for future business requirements.

  • Prototyping. This is like moving to a new and unfamiliar product or service. It’s a good way to challenge assumptions, but there’s always an extended learning time-frame. Prototyping is a good way to understand feasibility, applicability, and fit for achieving the desired outcome.

An experienced partner can reduce the risks of cloud migration, while meeting objectives at scale. With the constant expansion of service offers, there is a shortage of skilled cloud professionals. Dedicated IT specialists who have a proven track record of successfully implementing enterprise-scale cloud migrations can accelerate the process without increasing risk. 

If you would like to find out how to we can help you successfully migrate and/or engineer your IT infrastructure in the cloud, then give us a call or email us at

Other useful links:

How to plan for a successful cloud migration

Data & Analytics and Coforge

The State of Big Data in 2020

Let’s engage