Cloud technologies form the foundation of digital platforms in enterprises, with its faster, agile, automatable scaling characteristics. Enterprises start the Cloud adoption journey with a portfolio assessment, aiming to understand cost savings, with a set of quick wins, and experiments. In this process, the ‘elasticity’ of Cloud (on demand scaling of infra/ applications) does not make it in the priorities.
The two Cloud strategies – top-down and bottom-up – vary in accurately forecasting benefits of Cloud. The skewed expectation on top-down strategy is to model minute details of benefits, whereas the bottom-up strategy is prone to insufficient parameters to assess portfolio.
How to look at large portfolio for Cloud?
There are conflicting perspectives for portfolio assessment by infrastructure, application development managers and executive team.
The below table looks at these viewpoints, assuming thousands of applications and servers in the portfolio.
Viewpoints on portfolio applications and servers
Cloud vendors and service providers bring additional perspectives on the enterprise portfolio,based on their level of engagement. Viewpoints on portfolio applications and servers.
Vendor and service provider’s perspectives on enterprise cloud portfolio
These perspectives are distracting in a traditional assessment of portfolio
When companies execute ‘Application Portfolio Rationalization’, the objectives are beyond Cloud assessment, and does not address the divergent of viewpoints.
The approaches of performing traditional Cloud assessment are:
In particular, the ‘Server-centric’ approach (listed first in the table) is preferred by Enterprises because of its ‘measurability’; i.e. - the metrics and KPIs are easy to report. The infrastructure team, that owns ‘end of life’ programs, uses this approach with native skills and processes.
The parameters in server-centric assessment are:
The ‘Application-centric’ assessment calls for clarity of methodology and the KPIs are complex to measure. The application dependencies create complexity between servers, applications and other elements such as ownership. It is not uncommon to see orphaned applications in Enterprises.
The parameters of assessment of Applications are:
To contain risks, Automation is important, and traditional assessment covers it partially, with parameters such as:
The challenges of assessing impact points for application architecture include many definitions of complexity of the ‘as-is’ architecture (not the ‘to-be’ ones) conducted by disparate groups.
Other assessment methods such as Business Centric, PoC Centric, SaaS Vendor centric, and Cloud Provider centric contribute different outcomes to the overall assessment.
Need for a Newer Methodology
The divergence of viewpoints makes the Enterprises carry out multiple cycles of strategic evaluations of Cloud.
A newer methodology is required for stakeholders to set consistent priorities and this discussion suggests five consisting artefacts as standard.
The categories of stakeholders who need to be brought in a cohesive initiative are:
The initiative calls for not only revamping the traditional way of looking at the portfolio, but also to create five artefacts tailored to each organization.
Aretfacts 1: Benefits Classification models
Benefits Classification models (BCM) is an instrument to be invested upon by each enterprise, to consistently model and maximize beneifts across groups, with parameters and gudiance.
Severaltypes of BCM can be created;however,itis importantto have one model adopted across the enterprise.
Artefact 2: Reference Architectures
Finite set of Cloud Architecture Blueprints address the gap of assessing without architectures, and the blueprints evolve as the assessment progresses.
The distributed assessment teams will have reference-able architectures with consistentimpact points,reducing unwanted cycles of re-assessments.
Artefact 3: Tools Guidance
A coherent view of required Cloud (and migration) tools speeds up adoption and time-consuming certification of tools. The tools guidance refers to a standard mechanism (even self-service) for acquisition, qualification, provisioning, knowledge of tools.
The tools guidance is partly dependent on the reference architectures proposed for assessment.
Artefact 4: Integrated Consumption Model
For an effective Benefits Classification Model, a consumption model is required to understand the Cloud consumption patterns and practices across groups.This model addresses the varied set of Cloud providers and theirlicensing contracts that pose risks in measuring the benefits; in addition to the regulatory compliance risks and technical risks.
This discussion recommends a convergent enterprise platform than implementing platforms by vendors in the area of consumption monitoring.
Artefact 5: Assessment Methodology
A convergent methodology helps comparable categorization and inferences,to arrive at accurate outcome.
The challenges of having different perceptions on Cloud and its variations, added by the diverse methods of arriving at an adoption pattern, have not been addressed in a cohesive way in enterprises yet. The ‘convergence initiative’ and the methods discussed in this paper is an attempt to address these challenges. The convergence initiative looks at a newer way of assessing a portfolio of legacy applications, also looking at need for a more coherent Cloud directive.
Enterprises benefits from the methods discussed in this paper in ways of orchestrating the individual groups, practices and tools, and over and beyond, by enunciating a benefit articulation mechanism as well.