Save your runway: How startups can lower cloud spending

By Dmytro Grechko, founder and CEO at Deskree

Every software startup uses cloud services. They vary from simple hosting for front-end and databases to more complex services for container orchestration and network management.

If used right, cloud platforms help startups scale and develop their products fast and seamlessly. But if not, companies may face serious financial problems — even bankruptcy.

Using cloud services can run up costs

Multiple cloud providers, including AWS Startups, Google for Startups, Microsoft for Startups, offer early-stage startups free cloud credits. However, after they expire in 1-2 years, many of these companies may get a hefty credit card charge (surprise!) from their “big cloud brother.”

There are countless cases where the “surprise” was too big for the company to handle and it suddenly ran out of its own or investor money (or almost did). Startup Milkie Way receiving a $72,000 bill overnight for a test run on Google Cloud Platform is one of them.

Situations like these happen not because cloud providers want to rob startups. Instead, they offer documentation and a lot of guidance on reducing cloud spending. The actual reason lies in the misconception that every engineer knows how to work with the cloud.

In reality, very few engineers can properly set up and manage cloud infrastructure while optimizing the budget. They are usually known as DevOps and cloud architects. To avoid spending half of their runway on a cloud bill, founders should do two things.

The first one: Founders must ensure their company’s cloud architecture is a priority and handled by a professional. You don’t go to a vet if you have a sore throat simply because he is closer to your house. Similarly, your full-stack developer, regardless of their seniority, might not be the best specialist to work on your infrastructure.

And the second: To treat free cloud credits as if they don't exist. Startups shouldn’t strive to use up all of the credits, but rather to understand the real costs of running their software before they begin to pay for it.

How DevOps or cloud architects can reduce cloud spending

If you have someone on your team with experience in managing cloud infrastructure and who is fully committed to this job, they should prioritize the following tasks.

• Cloud cost analysis and monitoring. Start by conducting a thorough review of your cloud expenses. You can use built-in cloud cost management tools to monitor and analyze spending patterns or pick third-party solutions such as DataDog.
• Optimizing resources. Check your cloud setup to make sure your resources are effectively used. For example, you might not need a server running 24/7 if it only performs a cron job once a month. The same principle applies to disk space, memory, CPU, and GPU specifications. If you only use 10% of the database space you're paying for, you may opt for less space and set active monitoring and alerting to get notified when you reach a specific threshold.
• Using cost-saving options. Many cloud providers offer options such as Reserved Instances and Savings Plans to reduce costs.
• Using automated scaling and scheduling. This ensures that your infrastructure scales up during peak usage and scales down during periods of low demand. You can also schedule non-production instances to turn off during quiet hours, saving you money.
• Using serverless architecture. Consider adopting serverless computing models for certain workloads. Serverless architectures allow startups to pay only for the compute resources consumed during the execution of functions or tasks, so there’s no need to set up and manage dedicated servers.
• Optimizing data storage costs. Use different types of cloud storage depending on how often you access the data, implement rules for managing data throughout its lifespan, and consider archiving or deleting data that's no longer necessary. For example, Google’s Cloud Storage has different storage classes, and its costs, in particular, depend on the frequency of data usage.

How to reduce spending on your own

Not every startup needs a cloud specialist, especially in the early stages. Also, not every company can afford to hire one, as these positions are among the highest-paid in engineering. The base pay of a cloud architect is $100,000 a year and more.

If it’s your startup’s case, consider using tools that can help your team in orchestrating cloud infrastructure. For instance, if you’ve already chosen a tech stack and just need to deploy it quickly, consider trying Railway. This platform offers templates and user-friendly CI/CD processes for this purpose.

If your infrastructure mainly involves databases, storage, authentication, and simple APIs, tools like Supabase, Firebase, and AppWrite could be suitable options for building your MVP.

Also, you can try my company’s product — Deskree. It aims to make scalable cloud infrastructure accessible to any development team. One of the main reasons we developed this platform was to relieve early-stage companies of the burden of cloud engineering and associated cost optimization in product development.

Recently, we've also launched an Accelerate program where we aid startups in developing their APIs and architecture within days rather than months, and at a fraction of the cost.

Subscribe to our LinkedIn newsletter to get one useful article every week via email and on LinkedIn. Each story can be read in under 5 minutes, enabling you to promptly implement insights into your business.

More posts like this: