WHAT'S HOT
Back to the office?
Its funny. A few years ago, we used to ask how many days a week can you work from home. In the more progressive organizations, you would perhaps get the answer 'one day a week'. These days the question instead is, how many days do you work in the office?
What is your company policy? Most organizations we have surveyed work 3 or 4 days a week from the office. While people try to make those days common across colleagues, there seems to be no official rule. So often people only physically meet their colleagues once every two weeks. This works fine if you are experienced, and have known your colleagues well over many years.
Now imagine that you are a fresher in India. Just out of college, enthusiastic but raw. You need strong coaching and guidance - mentorship. Not only that, you need a strong sense of belonging. An understanding of the bigger purpose. This is much more difficult to do when most people work from home. Many companies in India have stuck to a three days from office schedule. Some have only two days a week. Why? Because people do not want to travel anymore for 3 hours a day. While it may work for the mid level or senior resources, it does not work for the junior staff - often 40% - 50% of the team. And this lack of training and mentorship is showing up.
In times of the pandemic, service delivery managers hired people from all over the country. Location was not important in a work from home environment. But now its an issue to get people together. In a recent delivery center visit by one client, they discovered that some people gathered there to meet the client, were meeting each other for the first time.
Then there are the clients who are wondering - if we only work in the office one day a week, why am I paying onsite rates for the supplier resources. What we see are some discussions to push more to offshore, or trying out nearshore locations for certain types of work - particularly agile delivery or new technologies. This situation is evolving and we are curious to see what direction its heading to.
No place like cloud
2023 has seen a continued move to the Cloud. More and more organizations (outside of financial services) made strong progress in moving a majority of their workloads into the Cloud. This shift is very evident in the infrastructure services contracts that we see in the market. Large datacenter contracts (50mn+) signed 5 years ago – are now a fraction of that value for the IT service providers (10-15mn), while rest of the revenues flow to the cloud providers. As Azure, AWS and Google cloud grow 20-30%, traditional datacenter management providers like Kyndryl, Atos, Fujitsu and DXC are faced with declining revenues. This has forced a shift in the industry with carve outs like IBM/Kyndryl and Atos/Evidian, and a re-focus from Datacenter services to Applications services / DevSecOps for companies like DXC and Evidian.
While there is a strong momentum towards the cloud, the cloud journey is not without its perils. A lot of the organizations who migrated to the cloud faced several challenges:
-
Lack of cloud specialist skills at both client side and service providers, made worse by large attrition figures at offshore centers. While all service providers talk cloud, we see a persistent lack of cloud skills. Often there are only a few cloud specialists supported by traditional infrastructure engineers.
-
Exploding cloud costs in the absence FinOps capabilities and lack of clearly identified responsibility for cost control.
-
Challenging integration of on-prem and cloud components – applications, network, etc., often with different teams working on each component creating support gaps.
-
Relative immaturity of DevOps/DevSecOps processes and ways of working across the industry, exacerbated by the lack of skills leading to operational instability and risk.
These challenges are clearly evident by a quick look at our squares. If we exclude the effect of End User services, satisfaction levels for Cloud services are markedly below those of Application services.
Our recommendations to clients embarking on a cloud journey are the following:
-
Do not underestimate the change and new skills required for your internal organization. Make sure to develop a strong internal capability (or Cloud CoE) for cloud strategy, cloud architecture, cloud governance and FinOps.
-
Perform a strong due diligence on the service providers' cloud capabilities and setup up clear operating principles, roles and responsibilities, and governance for the future operations.
The cloud manifesto
Now for some fun.
Question to ChatGPT:
Hello, can you provide me a list of challenges for companies moving their IT to the public cloud in the style of Karl Marx?
ChatGPT:
Certainly, here is a list of challenges that companies may face when moving their IT to the public cloud in the style of Karl Marx:
"The capitalist system imposes great challenges for companies moving their IT to the public cloud. The cloud represents a new form of technological infrastructure that replaces traditional forms of private ownership and control over the means of production. However, this shift to the cloud comes with its own set of contradictions and challenges.
Firstly, the reliance on cloud technology creates a new dependency on third-party providers who control and manage the cloud infrastructure. This dependency on a capitalist enterprise presents risks of exploitation and alienation of the workers involved in the production process.
Secondly, the implementation of the public cloud creates a new form of competition between companies, where the larger and more powerful corporations are able to monopolize the cloud infrastructure and dictate the terms of access and use to smaller and less powerful companies. This results in an unequal distribution of resources and opportunities, leading to greater economic inequality and class struggle.
Thirdly, the use of the public cloud creates new challenges for data privacy and security, as the control and management of data is transferred to third-party providers who may not prioritize the interests of the workers or users who generate and rely on that data.
Lastly, the implementation of the public cloud creates new challenges for the workers who are involved in the production process. The shift to the cloud may result in job losses or changes in job roles, as companies seek to automate and streamline their operations. This creates a new form of exploitation and alienation of labor, as workers become mere cogs in the machine of the cloud infrastructure, with little control or agency over the means of production.
In summary, the challenges presented by the implementation of the public cloud are rooted in the capitalist system of private ownership and control over the means of production. To overcome these challenges, a new socialist system must be established, where the means of production are collectively owned and democratically controlled by the workers themselves."
FinOps: Navigating the Art of
Financial Cloud Management
As enterprises harness features of cloud to deliver greater business agility, drive cost optimizations, and accelerate innovation, cloud adoption has gathered considerable traction. In 2022, cloud adoption by enterprises in the EU stood at 58% (as per Cloud Computing Statistics survey Sept 2023). A large number of these cloud migrations are Lift & Shift migrations. However, merely relocating workloads from datacentres to cloud doesn’t guarantee cost savings. Normally servers and databases are not optimized for high utilization in datacentres, a lift & shift move is likely to carry forward such inefficiencies into cloud.
Once in the Cloud, Enterprises quickly realize they are unable to meet the targeted business case. As cloud expenses increase cost management takes precedence, diverting IT attention away from cloud innovation. Streamlining cloud expenditure is a difficult task, for it requires combining both cloud and financial skills. Cloud engineers on their own, may lack the financial expertise to effectively design and operate a cost-effective cloud environment. Therefore, a collaborative effort drawing upon the knowledge across cloud operations, finance, IT & business leadership teams is needed to enforce the right behaviour. This has led to the creation of FinOps.
“FinOps converts the cost of workloads to its business value.
It positions you to manage cost overruns caused by suboptimal workloads
or misaligned pricing models. – Gartner”
Benefits of FinOps range from cloud usage understanding and costs optimization, performance benchmarking, real-time remediation, budgeting and planning.
We suggest considering the following key elements when designing your FinOps strategy:
1. Setup: FinOps function is cross-functional, and includes FinOps practitioner, cloud engineers/architect, and business. The FinOps function should form part of the CCoE. The size of FinOps team should be in proportion to the cloud setup. The utility of FinOps is measured by the ROI it drives and not the size.
2. Metrics: It is important to establish key KPIs and metrics that help enterprises measure cloud costs. Some of the prevalent metrics can be categorized into:
-
Visibility KPIs: These metrics create visibility across cloud environments (% resources tagged, cloud cost per application, cloud cost per user, etc.)
-
Utilization KPIs: These metric focus on utilization (Utilization rate of Reserved Instances vs Total Instances, % orphaned instances)
-
Optimization KPIs: To realize savings (Ratio of Optimized Cloud to Total Cloud Services, Hibernate % on weekends)
-
Automation KPIs: These KPIs relate to governance and automation (% automation changes that resulted in cost savings)
3. Cost Optimization: FinOps goes beyond tagging and account management. Intelligent decisions need to be made to reduce cloud costs. A mature FinOps setup will explore multiple cost optimization avenues.
-
Resource Optimization (Autoscaling, rightsizing, Usage of On-demand instances, Usage of Hybrid license).
-
Price Optimization (hyperscaler free credits, volume discounts (Reserved instances, Spot instances)
-
Architecture Optimization (Cloud Native /Serverless)
4. Tooling: With the right tooling, cloud consumption can be monitored and usage optimized by administering the right policies. Although all public clouds have native cost management capabilities, enterprises usually operate in Hybrid/Multi-cloud setup. In such situations it is better to leverage industry leading FinOps tools like Apptio Cloudability, IBM Turbonomics and Vmware CloudHealth. Good FinOps tools offer cost management, remediation, automation, and integration capabilities (with AWS, Azure, GCP, etc.).
5. Governance: Cloud offers speed - resources can be requested at the click of a button. The flip side to this the ability of developers to spin VMs at will is increasing cloud expenditure. Direct access to self-service in cloud must be governed through a service catalogue with necessary approvals. Governance must establish clear cloud usage policies. Policy alerts along with automated remediation need to be part of the governance. A good cloud governance setup encompasses cloud policies, guardrails and automation.
How to get organized?
Like any emerging discipline, there is shortage of FinOps practitioners globally. Lack of existing skills internally results in some enterprises relying on the managed service providers (MSP) to drive FinOps. This can be tricky as sometimes MSP service charges are linked to cloud consumption costs, and they don’t always have the authority to enforce the needed governance. We strongly recommend FinOps to be led internally. The direction setting and governance should be client enforced, even if execution is done by the MSP. Alternatively, specialized third party FinOps service providers are a good option.
Make sure to integrate FinOps into the initial cloud strategy (not an afterthought or add on). FinOps should be part of the CCoE, reporting into the CIO or CFO (or both). The FinOps team should have the mandate and authority to enforce governance. Provide fundamental cloud training to all key stakeholders, as the effectiveness of FinOps relies on the cooperative effort of all essential cloud stakeholders. When implemented well, FinOps can delivery cost optimization in the range of 20-30%.
Success!
![process_optimization_3 [Converted]3.jpg](https://static.wixstatic.com/media/4941a0_cdd4db76bd9d45d19b3f3c799f69ec08~mv2.jpg/v1/fill/w_980,h_560,al_c,q_85,usm_0.66_1.00_0.01,enc_auto/4941a0_cdd4db76bd9d45d19b3f3c799f69ec08~mv2.jpg)
Service Integration: Guiding the ICT function to success in today’s multi-vendor environment.
As enterprises continue to expand their digital initiatives, newer technologies continue to be added onto the existing ICT portfolio. Rapid proliferation of cloud, AI, IoT is creating a demand for newer skills and making technology sourcing complex. As newer technologies integrate into the portfolio, specialized providers begin to operate alongside partners, crowding multi-vendor ecosystem.
Although multi-vendor strategy offers the advantage of flexibility and scale, it does have some challenges like:
-
Fragmentation of IT ecosystem: Service providers start to operate within their own frameworks, focus on their own SLAs leading to siloed delivery.
-
Inconsistent Service Experience: Each service provider has his own distinct flavours of process and tools. Industrialized delivery models with factory like setup results in inconsistent service experience across service providers.
-
Overlapping scope across vendors: IT process automation like E2E monitoring, RPA and Chatbots cuts across service provider towers. If each service providers deploys his own IP for automation there will be duplication of tools and effort.
-
Lack of end-to-end visibility: In the absence of strong governance information sharing across providers is uneven leading to poor compliance and end-to-end visibility.
“76% of ICT departments do not have the needed capabilities to manage
and steer their service providers towards success.”
Partnership Benchmark findings
The typical ICT teams are filled with techno-functional experts who have sound technical skills. They have always been do’ers and could be counted to deliver complex technology solutions. However, in a multi-vendor environment they are now thrusted to steer vendors, a skill that they need to evolve.
How does ICT then steer a group of service providers to deliver end-to-end service? How do they ensure that the service experience towards business is consistent? How do they ensure information exchange and collaboration across providers is seamless? How do they deliver IT transformation projects cutting across provider ecosystem?
The onus is now on the CIO to lead the transition of the ICT teams from a group of “do’ers” to “managers” equipped to manage eco-system partners. The CIO must start by setting up strong vendor and service management capabilities within his organization. SIAM (Service Integration and Management) can be used as the framework to drive this change. The aim is to get all parties to collaborate to deliver a seamless service by working “as one team”. By laying strong SIAM foundations, ICT deliver efficient service management by orchestrating strong governance, and collaboration between all parties.
Some of the components for an effective Service Integration function:
-
Multi-vendor Collaboration: A multi-tier governance framework for cross vendor collaboration designed to foster spirit of collaboration, spur blameless culture, and seed continuous learnings.
-
E2E Service & User Experience: E2E SLAs encouraging service providers towards E2E outcomes. Service outcomes like XLAs (Experience Level Agreements) and business KPIs define these outcomes.
-
Single Source of Truth: Common integrated tooling and processes delivering E2E visibility. This is particularly true for the ITSM tooling and related processes (end to end ticket management, change management, etc.)
-
Culture of Ownership: Driving Organisation Change Management (OCM) within the ICT teams helps re-focusing of the internal IT staff, such that they are better able to control and steer service providers outcomes.
Service Integration function can be run in multiple ways.
-
Internally Managed: ICT themselves take up the SIAM role. This will need some initial effort to re-focus internal teams towards vendor management.
-
Dominant Service Provider: Here one of the service provider (typically delivering one or more service towers) additionally manages (parts of) the SIAM function.
-
Specialist Service Integrator: a third party (e.g. S-Square) will run the SIAM function. Suitable in cases where ICT bandwidth and capabilities are limited. Specialist consultants can be called upon to set-up and stabilize the SIAM and OCM.
With many years of experience in the field of IT outsourcing, we are a firm believer that the success of the SIAM function lies within the internal ICT organization. Irrespective of how SIAM is setup, ICT still has a vital role in managing the service providers. ICT should be setting up the governance to steer and challenge the service providers, they should facilitate seamless collaboration across service provider on both BAU (Business as Usual Services) and IT Transformations.