3 Ways to Reduce Cloud Waste and Improve Cloud Efficiency

Cloud spending is on the rise, but much of it drives little business value. Here’s how to identify wasteful cloud spending and reduce your cloud budget.

Written by Phil Pergola
Published on May. 21, 2024
ceo grabbing dollar bills falling from cloud illustration
Image: Shutterstock / Built In
Brand Studio Logo

Cloud spend is on the rise, and with it, cloud waste.

Common estimates suggest that about 32 percent of cloud spend is wasted, and these days, that extends beyond traditional conceptions of the cloud. Alongside AWS, GCP and Azure are specialized vendors like Snowflake, Datadog and Databricks, as well as everyone’s favorite new talking points, AI and generative AI. As AI adoption rises, so do costs: Large language models require a huge amount of computing power and cloud spend, as the cloud is an essential technology enabling AI.

3 Tips to Reduce Cloud Waste

  1. Don’t pay for cloud services you don’t use.
  2. Set up targeted, timely spending alerts.
  3. Calculate and manage unit costs.

But the cloud’s integral role in powering innovation causes many companies to use it indiscriminately. Cloud waste accrues in the form of unused resources and services, inefficient cloud architecture and unclear unit economics. Companies outlay huge sums of money that drive little to no business value.

Let’s look at three methods SaaS companies can use to reduce their cloud costs and boost their overall cloud efficiency.

 

1. Don’t Pay for Cloud Services You Don’t Use

The cheapest cloud resource isn’t the most heavily discounted one — it’s the one you don’t pay for in the first place.

This may sound obvious, but most (if not all) organizations are paying for cloud resources or services that aren’t driving any business value. It’s not because they’re negligent, it’s because the complexity of cloud spending — multiple clouds, multiple vendors, shared spend and containerized spend — make it very difficult to understand the business impact of every cloud resource.

In order to stop paying for valueless resources, the first thing organizations have to do is quantify the business value of each cloud resource. You’re a digital workout company: How is your shared S3 bucket, a public cloud storage container, impacting the profitability of your business? You’re a cybersecurity company: What return are you getting on your CloudTrail investment?

To answer these questions, the first thing businesses need is a complete cost allocation. This means tracing every penny of your cloud spend to the appropriate sources, or dimensions, be they customers, products, features, microservices, teams or whatever else is important to your business. With accurate dimension costs, you can calculate their profitability and make informed choices about where to refine.

For context, less than half of organizations feel confident doing this, according to a CloudZero’s The State of Cloud Cost in 2024 survey, offering a strong indicator of why companies feel such acute cost reduction pressure.

Allocation also spans attribution: Assigning specific costs to the engineers responsible for that cloud infrastructure. Once engineers have a clear idea of what costs they’re responsible for, they can decide which resources and/or services to eliminate. In the span of a single call, I’ve witnessed engineering teams use this visibility to eliminate hundreds of thousands of dollars of annualized spend.

More on the Cloud3 Tips for Successful Cloud Adoption

 

2. Set up Targeted, Timely Spending Alerts

Many cost platforms focus on reactive cloud cost reduction, or cutting costs after you’ve already overspent. But for true cloud efficiency, you need proactive cost reduction: keeping costs low before they spiral out of control.

One powerful way to do this is to set up targeted, timely spending alerts, also known as anomaly alerts. In cloud spending, an “anomaly” is a period of unusually high (or low) costs within a particular resource. An anomaly alert is an automated message sent from your cloud cost management platform to the engineer(s) responsible for it.

In anomaly alerts, there’s a gold standard to look for:

  • Hourly granularity: Your cost management platform should be able to pinpoint the exact hours of the anomaly. Day-level granularity omits anomalies that happened within a particular window.
  • Team-specific alerting: The alerts should go only to the teams/engineers responsible for the affected infrastructure. Why? Because if engineers get too many irrelevant notifications, they stop paying attention. This gets back to allocation and attribution. If you’ve allocated and attributed 100 percent of your cloud costs, you can be sure that all notifications are relevant.
  • Savings tracker: Your cloud cost optimization platform should track all the savings you’ve realized centrally. This is both so that you can quantify the cost efficiency of individual dimensions and to power a positive feedback loop within engineering teams. When they can see the concrete impact of their work, engineers are more likely to address new anomalies.

Anomaly alerts drive engineering ownership of cloud costs. CloudZero’s survey showed that this leads to better business outcomes, as 81 percent of respondents said cloud costs “are about where they should be” when engineering has some level of ownership.

 

3. Calculate and Manage Unit Costs

Eliminating waste doesn’t just mean giving your cloud bill a haircut every so often. It also means scaling efficiently. As your organization grows, your cloud costs will naturally grow, too, just like headcount, office space or any other core business investment.

The trick is to make sure your cloud costs are growing efficiently. How do you do this? By tracking core unit cost metrics, which 66 percent of companies struggle to do. Your cloud costs may grow, but if your unit costs stay flat or trend down over time, you’re preventing and/or eliminating cloud waste.

If you’re a digital workout company, tracking the cost per workout will tell you how much it costs you to keep your customers fit. If you’re a cybersecurity company, tracking your cost per firewall will tell you how much it costs you to keep your customers safe.

If unit costs are trending down, excellent. If they’re trending up, what do you do? Go back to steps one and two: Figure out which dimensions are driving the increased costs, who on your team is responsible for managing them and where they can drive efficiency.

More on the CloudTop Security Risks of Cloud Computing

 

Maximize the Return on Your Cloud Investment

Understandably, many companies face pressure to reduce costs, and there are some low-effort ways to do so without hurting business value. But ultimately, managing cloud spending is about maximizing return on investment: getting as much bang for every cloud-native buck you spend. Allocation, attribution, anomaly alerts and unit cost tracking form a comprehensive mechanism for maximizing your cloud ROI.

Explore Job Matches.