Why is our cloud infra bill so high? Does it need to be?
Tech leaders have a lot on their mind right now and the hidden costs of cloud may not be a the top of the list. Seeing your LinkedIn feed constantly flashing mass layoffs left and right, budgets getting cut, economy slow downs… it’s all very depressing and hard to avoid. Every tech company will feel the pinch in some way or another. Cloud services bills have a tendency to run away like a dog chasing a fresh bone. When the sun is shining you just do whatever you need to do to ensure that your infrastructure stays up under the immense weight of customer traffic and orders etc.
But when the economy turns a corner, orders are down, revenue starts to dip – the big $$ on the cloud bill start to stand out like a sore thumb. Now, understanding the hidden costs within your cloud becomes critical and knowing how to optimize these costs while maintaining services for your customers.
Next thing you know you’ll have an invite for a zoom call to “review current <insert cloud provider> operating costs”.
To gear yourself up for that chat, do some homework to understand where your cloud costs are going. With this knowledge you’ll be better prepared for that chat and know exactly why your spend is the way it is. And why it should (for the most part) stay that way.
I. The Hidden Cost of Cloud Compute
One of the greatest benefits of cloud services is the infinite flexibility when it comes to scaling. Being able to scale up the compute power for your applications at the drop of a dime is invaluable to a business that relies on high uptime and availability of services to their customers. For large retailers the cost of downtime can be astronomical. Ranging anywhere from $8,000 – $25,000 an hour all the way up to $5 million an hour for the largest of enterprise businesses. With this in mind, it’s no wonder cloud infrastructure teams will often overprovision their compute resources to avoid this potentially catastrophic issue.
But when these high levels of compute aren’t being used, it can rack up a hefty bill during quiet periods. Left alone, your return on investment (ROI) is slowly rotting away and your operational expenditure is steadily climbing. During economic downturns, this is the last thing you want on your balance sheet.
Luckily companies like AWS have put in place ways to optimize your compute costs to ensure that you’re only spending what you need and when you need it.
Here are a few options that AWS offers that are worth considering:
- Spot instances
- Reserved instances and savings plans
- Auto-scaling
- AWS cost explorer
- AWS compute optimizer
II. The Hidden Cost of Cloud Storage
Storage is another area where costs can easily creep up as data storage needs grow organically with your business. With the amount of data that is being captured by businesses today to fuel innovations like machine learning and AI, transactions for eCommerce businesses, and marketing data – this all needs to be stored somewhere.
When it comes to storage, the cost boils down to you asking yourself these key questions:
- What level of durability do you need?
- How available does your data need to be?
- How long do you need to keep the data for?
- How quickly do need that data when you want it?
Having clear answers to these questions will help you take advantage of a number of options to make savings. Think to yourself, how will these impact my core business? What risk does this introduce? and how tolerant is my business of this risk?
For example, AWS has excellent options available if you have clear data policies on how you want to handle your data. Using products such as S3 Glacier Deep Archive can save you an immeasurable amount of cloud costs if you need to store data for compliance purposes, but don’t need immediate access to that data.
III. The Hidden Cost of Bandwidth
When it comes to “hidden costs” this is an area that is often neglected. The vast majority know you need to scale up compute and storage to mitigate risks and ensure availability. But where many will fall short is in accounting for the costs of transferring data. Where data transfer costs can really pile up is with outbound data. For example AWS EC2 egress costs anywhere from $0.05 per GB for over 150TB to $0.09 per GB for the first TB!
But there are ways to limit outbound data, such as:
- Using a CDN such as CloudFront to reduce the amount of outbound data from your compute instances to the public internet.
- Transferring data mainly within a single region. Keeping common data transfers between services such as EC2 and S3 within a single region will greatly reduce costs.
- Not all regions have the same charges for data transfer. So when architecting your solution, ensure to take into consideration the costs of scaling within your chosen regions.
Equip Yourself with a Plan
Don’t get caught out when the budget and cost savings talks start at your company. With just a few hours of digging and some proposed tweaks to your cloud architecture, you can unveil the hidden costs of your cloud and make quick changes that will have an immediate impact that will compound over time.
Having a firm grasp on your cloud costs will instill confidence in your area of the business and ensure that your area stays as well-funded as possible.
Getting a handle on cloud costs is no small feat. You and your team may be more than capable of tackling this in preparation for the coming financial year planning. But, if your team has other priorities and this needs to be done – don’t hesitate to reach out to us. We live and breathe this. Helping growing businesses leverage cloud services to the best of their capabilities is our mission.
