Why are cloud costs so high?

If developers can’t control their Cloud spend, then the product owners can’t control the gross margin. Developer platforms must enable the right cost control and accountability to get the right outcomes for the business.

Time to read
9 minutes
June 19, 2024
Jon Shanks

As companies grow their Cloud footprint, they also significantly increase their operational costs. With the evolution of Cloud and ever increasing number of services that they offer, it is becoming more apparent that many organisations have yet to master the expenditure and sustainability controls they need.

Key Takeaways

Average Cloud wastage is 27%, estimated to be $186.38bn in overspending globally.

The Cloud now has a greater carbon footprint than the airline industry.

Utilising Cloud best practices will reduce the amount of engineering the platform team has to do and provide better scale and reliability.

With Cloud revenue projected to reach $630.3bn this year, (2024), the number of organisations accepting it as part of their growth is evident. So, how do companies stop themselves from contributing to the 27% Public cloud wastage or in financial terms, adding to the whopping $186.3bn financial waste that is set to happen this year?

This article will cover six of the most common constraints and hopefully provide you with some key adjustments your organisation can make. A follow-up article  will provide more detail on what you can do to get more control and become a more frugal and sustainable business.

Why are businesses spending more than they need to on the cloud?

Several factors contribute to unnecessarily high costs being encountered when consuming the Cloud, these usually fall into one of the 6 categories below:

  1. Predictive costs: teams are not aware of the projected hosting cost of their service until it is deployed which costs teams money
  2. Realtime cost insights: product teams often have no ongoing real-time cost transparency that helps teams become more cost-conscious
  3. A lack of architectural setup and planning: teams start to consume the cloud without the right architecture in place to make budget, costs, commitment discounts and transparency simpler
  4. Poor cloud-native app architectures: applications are not being designed appropriately for the Cloud and lack modernisation
  5. Product teams lack cloud control: developers in product teams don’t have the control they need to adjust infrastructure, right size and delete unused resources. 
  6. Costs are not part of delivery: Product teams are not making costs part of their delivery processes

Predictive costs

Due to the complexity of the Cloud and the Cloud costing models, there is a big disconnect between product owners, developers and product managers in understanding how much their service will cost when hosted in the Cloud. 

With various pricing models, discounts and service offerings it makes it increasingly difficult to have a good understanding of costs before they have been incurred. Without a solution in place that can simplify this, teams are unable to investigate alternative cost-saving initiatives or be able to report a more accurate hosting budget back to the business. 

Traditional operating models such as DevOps, Platform or Cloud engineers brokering Cloud delivery on behalf of teams, is creating a layer of abstraction between product teams and the Cloud. This unfortunately complicates matters by developers not being in control of costs and the “Brokers” not being accountable. 

Real-time cost insights

Real-time cost insights provide clear visibility into the associated costs of the resources being provisioned in the Cloud. It creates a proactivity in the team that will encourage people to attempt to save money where possible. 

Teams can often:

  1. Have too much information i.e. have visibility of all cloud costs with little context to their own resources
  2. Have too little to no information on the services, their cost and the appropriation of how those cloud resources relate to the application

This is often down to either a lack of good tagging structures that can get granular or a lack of isolation of the Cloud resources or both. 

Without clear insights, teams will be unable to look into applying unit economics as a key metric to gauge the efficiency of spending in the cloud. For more information on Unit economics in the Cloud please read the FinOps introduction.

Architectural setup and planning

All major Cloud providers such as AWS, GCP and Azure offer a set of architectural principles that enable cost allocations, budgets, discounts, alerts and visibility. Although every provider has a different level of maturity around these services, the fundamentals are pretty similar:

  1. Isolate workloads to their teams to make cost allocations simpler and easier to understand
  2. Define a tagging policy that covers all the key business information required for ease of allocation and understanding
  3. Enforce tagging policies using security policies, so resources are unable to be deployed with the tagging standards in place
  4. Set budgets for cost, usage and reservations
  5. Set alerts and notifications to the necessary teams so they have an awareness of any increased spend
  6. Use grouping mechanisms to scope budget policies appropriately to the relevant teams
  7. Set limits using security policies to prevent budgets from being exceeded if appropriate, (this could cause operational issues with scale demands)

Without these in place, it will be increasingly difficult to have the right level of cost isolation, granular understanding and control. Although 51% of organisations have FinOps teams in place, there is still 27% being wasted on the Cloud. With nearly a third of organisations spending more than $12 million a year in the cloud, that’s $3.24 million a year being wasted per organisation. 

This is often due to the heavy reliance on third-party products, as opposed to adopting the recommended Cloud Landing Zone and well-architected frameworks of the cloud providers. 

Cloud-native application architectures 

With the largest revenue earners of the Cloud still being virtual machines, it’s clear that businesses are not modernising their applications to leverage more cost-effective cloud services. 

With the view that modernisation will follow a migration, a lot of the modernisation work gets deprioritised due to unforeseen migration challenges. 37% of cloud migrations fail to meet their objective of making the cloud an integral part of their core strategy due to over-running, poor stakeholder engagement and migration approach. 

Pro-longed migrations have the added disadvantage that workloads are in a state of limbo for a prolonged period, therefore unnecessarily increasing cloud costs. A modernising strategy should be part of the migration strategy so that services can leverage the Cloud appropriately and reduce costs. 

Product teams lack control

As mentioned briefly in other areas of this article; Developers in the product team are often abstracted from the Cloud, as they have no direct control over the infrastructure they are inhibited from impacting costs. 

With many developers lacking a deep understanding of the Cloud, specialist Cloud engineers such as DevOps or Platform Engineers facilitate infrastructure provisioning on their behalf. This means the infrastructure for application environments is created independently and then leveraged inside of application deployments. 

This has some key disadvantages: 

  1. The product team has no visibility on how much things are going to cost, (estimated costs) to then make the appropriate adjustments
  2. They have no direct control over the type of infrastructure or the resource limits such as CPU, memory and disk, meaning it can often be more than they require
  3. They don’t control the duration of the infrastructure, meaning it is often idle until it is needed again as opposed to being deleted
  4. They don’t control the application dependency infrastructure such as databases and message queues, resulting in unused instances as requirements changed or underutilised resources that they are unable to modify
  5. They can’t delete things they are no longer using or require

With the ability for teams to make the necessary changes needed to impact the costs, product investments will be able to yield a suitable return on investment. Understanding the cost is only one part of the solution, there also has to be a way for teams to modify the infrastructure. With developers having the most knowledge of the application, environments, dependencies and peak usage times, they are the best people in the business to take control of the changes. 

Cost planning with product delivery

Although many product teams often budget as part of their overall delivery, this is usually a rough estimation and is often not granular, accurate or happening with enough frequency. 

Despite efforts to manage cloud costs, 78% of companies allocate less than 75% of their cloud spend accurately, this means that accurate costs still need to be improved for many organisations. 

Without the right level of accuracy and insights, teams will be unable to adopt a unit economics approach to the Cloud. This means teams won’t be able to accurately show how profitable the product is or how soon it will achieve profitability without making costs part of their everyday feature delivery and implementing a unit economics approach to the Cloud. 


Managing cloud costs remains a significant challenge for many businesses, often leading to unnecessary financial waste and increased environmental impact. With public cloud revenue projected to reach $690.3 billion in 2024, organisations need to take proactive measures to control their cloud expenses effectively.

Key constraints contributing to high cloud costs include a lack of predictive cost insights, insufficient real-time cost transparency, inadequate architectural setup and planning, outdated cloud-native application architectures, limited control for product teams, and costs not being integrated into delivery processes. Addressing these issues requires a comprehensive approach that involves implementing best practices, adopting proper tagging policies, setting budgets and alerts, and empowering product teams with the tools and knowledge needed to manage their cloud infrastructure directly.

By focusing on these areas, organisations can reduce wastage, improve cost efficiency, and contribute to sustainability goals. Companies must view cost management as an integral part of their cloud strategy, ensuring that the money spent is justified and contributes to the overall profitability and sustainability of their business.


Related Posts

Related Resources