During a recent Apptio teleconference that reviewed services and discussed cloud adoption in the financial services industry, Leandro Santos and Himanchu Agarwal of McKinsey & Company shared these 10 core beliefs on cloud transformation:
- Public cloud will dominate the cloud market in 2020
Organizations could drive significant business value by quickly leveraging public cloud vs. private cloud, and aspire to migrate as much as 80% of workloads onto off-premise cloud. In the short-term the migration scope could be as much as 30-50% depending upon the workload characteristics (e.g., mainframe, latency requirements, regulatory and security needs).
- Maintain 2-3 cloud providers to avoid long-term vendor lock-in
They can achieve this by leveraging off-the-shelf commercial solutions (e.g., Bracket Computing, Docker) to drive standardization and inter-operability.
- Most legacy cost elements (e.g., software, hardware maintenance, facilities, labor), are ‘shared, sticky and invisible’
They need to be ring-fenced and proactively re-engineered to rationalize costs.
- Target state consumption of non-labor is typically 20-40% lower than in traditional data center
This is driven by having a usage-based consumption model, and rationalization of DR capacity and duplication copies. Additionally, more than 50% of the total cloud benefits can be achieved without code-level application remediation; many organizations aspire for cloud-native applications which can be cost-prohibitive and time-consuming.
- In public cloud, infrastructure is consumed as a software
Organizations need to be structured as agile teams to:
a) build/automate new services, and
b) manage ITIL-like processes e.g., incident management.
- Cloud requires a small (less than 100 FTEs) but highly skilled labor force
Building and retaining the team requires a radical approach (e.g., training with the cloud providers, incentives aligned with impact, fast-growth career track). Further application development and a support model can be evolved to enable DevOps capabilities (e.g., self-service, 1-click provisioning) to unlock 20-30% of additional productivity.
- Legacy migration should adhere to the defined standards to retain the automated operating model
Environment blueprinting is the most effective path, especially for applications requiring frequent touchpoints with Infra groups. Cloud migration schedules need to be architected to optimize commercials (e.g., align migration with application refresh cycle and other on-going initiatives, moving DB farm to the cloud as is to maintain license costs).
- System integrator capabilities can be crucial
They are crucial to building the migration factory but contracts should be set up to pay on value realization. Migration factory capabilities can be retained in situations where M&A is a strategic business priority
- Focus the change story
The change story should be focused on the business impact of cloud transformation vs. technologies being deployed.
- Demonstrate material business impact
Do this on a frequent basis (e.g., every 3-4 months) to sustain momentum e.g., off sites with key stakeholders.
Questions on McKinsey's core beliefs about cloud transformation? Please email Leandro or Himanshu with any questions. Or even better yet, join us for our next Teleconference on July 19th when McKinsey will return to discuss these insights on cloud transformation and more.