7 Predictions for Cloud in 2019

Frédéric Cruchet, VP, Global Head of Cloud Services at Sogeti discusses his top seven cloud predictions for 2019.


1. Migration to cloud becomes mainstream

With cloud native applications proving a versatile tool for speeding up and driving down costs in customer-facing/front office activities, it is now the turn of the back office to benefit from cloud capabilities. Moving legacy to (or integrating it with) the cloud will deliver the speed, flexibility and cost benefits already being seen in the front office. The objective being to move from today's 2-speed IT (legacy v digital) towards a more balanced IT landscape. While not all applications will run on cloud platforms, 2019 will be the year in which cloud goes mainstream.

2. Provisioning as code marks a new era for developers

2019 will see an increase in provisioning as code, notably infrastructure as code, configuration as code, and deployment (pipeline) as code. This is about providing developers with scripts they can use to automatically build infrastructure architecture, define configuration, and finally deploy CI/CD pipeline to enable DevOps way of working. We will see developers working faster and doing more due to automated provisioning of infrastructure that reduces manual effort. However, this brings the challenge of how best to manage all the scripts. It will be important to optimize reuse of existing scripts and quality development of new ones. Management of a scripts library will be a key differentiator for better efficiency and agility of developers.

3. The use of multi-sourcing strategies will increase

A focus on cost optimization in sourcing will lead to more multi-sourcing strategies as enterprises move to multi-cloud provisioning. Previous concerns about security in the public cloud have been alleviated, with assurances that both public and private are equally secure. One driver for greater public cloud adoption is that it is a commodity market and therefore less expensive than private cloud. We expect to see more clients combining public cloud offerings from Microsoft, Amazon and Google in an ecosystem of vendors. But organizations will need support in managing this complexity, sourcing the right MSPs and longer term management of the cloud ecosystem.

4. Public versus private (on premise) still needs to be managed

While cloud will become mainstream, some enterprises will always have legacy on-premise solutions. With a 100% public cloud model being unlikely for most of our enterprise customers, a hybrid approach to managing the cloud will be needed for the foreseeable future.

5. Cloud native is a focus for new business model development

In 2017 Capgemini and Sogeti produced the 'Cloud native comes of age' report. At that time, one-sixth (15%) of respondent firms' stated new applications were built in a cloud native environment. Fast forward to 2019 and I predict this figure to grow significantly. In particular, organizations with omni-channel strategies (i.e. Retail, Financial Services etc.) will increase the use of cloud native applications in their engagement with customers via smartphones, tablets etc.

6. Containers become the default landing platform

Containers are the new IaaS (Infractructure as a Service). For me, containerization such as that offered by the likes of Kubernetes is ideal for the rapid development of applications in the cloud. Containers automate release management, with all the different components managed within a single container which, I predict, will become the default landing platform. Going serverless will then be the next evolution, but perhaps we'll need another year before this reaches a maturity that suggests more widespread adoption.

7. Cloud to the edge deployment will be needed to manage IoT (sensors)

Expect to see a shift towards edge deployment (both provisioning and continuous integration/continuous development) for the cloud. Why? To manage the proliferation of data from the phenomenal growth of sensors and other IoT devices. With this growth there could be an issue with network connectivity and latency. A move to the edge (provisioning close to the sensor, instead of in a centralized datacenter) is a way to manage some of the data volumes directly on the edge. This will be also balanced with deployment of 5G.


If you would like to discuss any of the above, get in touch and we'd be happy to answer any questions you may have.