Optimizing AWS Spend as a Telecommunications Company
Just because telecommunications is an industry known for its complexity doesn’t mean communication service providers can’t find straightforward ways to optimize operational expenses.
In early 2023, technology research firm Omdia looked at the financials of 28 major telecommunications (telecom) companies, going back a decade. It found that only nine of those companies had reduced their operational expenditures in that time, and only three of those nine had done so while keeping their ratios of operating expenses (OpEx) to revenue below 60% (a sign of efficiency).
At Slalom, we think it’s no coincidence that few of the telecom companies Omdia surveyed had many technology systems running in the cloud. But even telecom companies that have migrated large sets of workloads to the public cloud are facing challenges managing their OpEx. We’ve been working with an increasing number of telecom companies to control their cloud spend and deal with, in Omdia’s words, all the “necessary shifts toward software-centric networks and the delivery of new digital services” and “the sheer complexity of maintaining and managing heterogenous infrastructure.”
Our work with telecom customers to optimize their OpEx affirms the value of starting small and the effectiveness of best practices — two in particular.
The power of a proof of concept
When cost optimization is your priority, testing a new service or architecture on a small portion of your application portfolio is a low-risk approach to exploring savings opportunities on a much larger scale. For example, we recently worked with a Fortune 100 telecom company that was considering moving its applications — which were already running on Amazon Web Services (AWS) — to AWS Graviton2.
- Start small: We surveyed 16 application candidates for the migration but chose one to try out based on its size, maturity, and other key factors.
- Test expertly: We built the application on Graviton2 and implemented AWS best practices including right sizing and auto scaling.
- Scale success: Based on the application’s performance, we were able to estimate that a portfolio-wide migration could save the company 82% on compute costs with the help of right sizing and $3.8 million annually with the help of auto scaling.
Put best practices to work
New to right sizing and auto scaling? These are essential areas of exploration for telecom technology teams eager to optimize cloud costs, including AWS spend.
Right sizing
AWS defines right sizing as, “the process of matching instance types and sizes to your workload performance and capacity requirements at the lowest possible cost.” Components of telecom business support systems often have different instance type requirements — i.e., one size does not fit all — and yet AWS observes (as do we) that many telecom companies tend to bring an on-premises mindset with them to the cloud by overprovisioning resources, or “oversizing” their instances, because they want to be sure they can handle peak usage. Right sizing your instances allows your organization to handle extreme traffic spikes without overprovisioning, especially when combined with auto scaling.
To make a best practice even better, it’s possible to automate right sizing and not enough organizations take advantage of this. According to Flexera’s 2023 State of the Cloud Report, less than half of organizations practice right sizing, with 44% using manual policies to do so and just 33% using automated policies.
Auto scaling
In the telecom industry especially, optimizing cloud spend is frequently a matter of scaling down as well as up, and in as well as out. Right sizing is a form of vertical scaling. Auto scaling is a form of horizontal scaling, or “elasticity.” As PC Magazine explains it, “When traffic spikes for a certain period of time, auto scaling creates a new virtual machine (VM) instance to manage it and disengages the VM when the traffic decreases.”
Auto scaling was crucial for our telecom customer to achieve the $3.8 million in savings that we estimated it would by moving to Graviton2. Significant savings for telecom companies using auto scaling aren’t unusual. One AWS case study examines Australia-based mobile virtual network operator amaysim’s use of AWS Auto Scaling and diverse instance types to save 75% on compute costs.
Don’t forget the most important part
Controlling OpEx requires operational change, and for any big organizational change to take hold, you need people and teams to be change agents. Omdia cites training and cultural shifts as requirements for sustained cost optimization and we couldn’t agree more. That’s why we recommend that organizations interested in optimizing cloud spend consider cloud financial management, or CFM.
A 2022 report from AWS defines CFM as, “a set of principles and practices for infusing cost awareness and accountability into the provisioning, deployment, and monitoring of cloud resources.” If it seems like CFM is just about saving money, then think again. CFM is about allowing your organization to unlock value by fully taking advantage of the variable costs of the cloud. From a people perspective, we think it’s also about bringing together finance, business, and technology teams to think holistically about cloud investments so they can develop effective ways to manage cloud spend and measure value.
Want to learn more?
We have a team of cloud experts who are certified by AWS to help organizations assess where they are in their CFM journey. Learn more by downloading our guide to CFM or reach out to us about getting started with a CFM workshop.
Slalom is a global consulting firm that helps people and organizations dream bigger, move faster, and build better tomorrows for all. Learn more and reach out today.
