Easy tips to control cloud spend

Too often we hear the same issues being raised by procurement and finance departments across the land “why is our bill spiralling”, “who authorised this”, “which department bought this service”. 

The upsides of cloud are well known, releasing the innovation inside your business, transforming how you deliver services and letting you reach your end-users faster and in more creative ways. 

The downsides however don’t often make the headlines. Too often we hear the same issues being raised by procurement and finance departments across the land “why is our bill spiralling”, “who authorised this”, “which department bought this service”. In our experience nearly every enterprise cloud consumer is wasting between 20% to 70% of their cloud spend, with the average being more than 35% in a typical mid-sized IT environment.

Here’s a quick top-10 of the biggest areas of cloud wastage that we’ve seen over the last 5-years:

1. “Lift and shift” when migrating to the cloud - In an ideal world, you would optimise before migrating to a new cloud platform – but not every organisation has the luxury of endless time or dual running budgets. Too often we see companies taking their existing over-sized and under-optimised platforms and services and transitioning them to their new cloud provider. Not only is this technically rarely the right option, but commercially it can more than double your cloud spend, in some cases increasing spend by 4-6x.

2. Unused & Inactive Instances - In the same way the cloud makes it easy to spin up compute resources, it’s just as easy to forget about them. Unfortunately, you pay for your cloud infrastructure whether you use it or not, cloud sprawl with orphaned instances created by shadow or development projects are often left sitting idle for months or years, racking up bills day by day.

3. Instance Over-sizing - This one’s a bigger culprit than you might expect! In the cloud era, too many developers still have a legacy mindset when sizing compute resources, provisioning far more capacity than is actually required, ignoring right-sizing, avoiding auto-scaling and taking a traditional approach in an agile world.

4. Failing to Review Pricing Options - It’s critical that you stay on top of your cloud providers pricing options as these change on an increasingly frequent basis, not only with new offerings but also with new commercial models. Every cloud provider has flexible pricing options that allow you to make an upfront commitment in return for a discount on your compute cost. Understanding your usage patterns is invaluable, knowing what resources require true flexibility and which are more persistent. Doing so will arm you with the knowledge you need to decide between instance sizes, contract terms and importantly how to use RI’s instead of “on-demand” resources or increasingly where you can leverage the exceptional low prices of Spot Instances for more of your workloads. You need to architect for these changes (particularly for Spot), but once you've done so you can easily achieve savings of up to 80% of your monthly bill.  It's a great way to way to buy much bigger instances for a fraction of the usual cost.

5. Storage & Data Protection - Just like with on-premise platforms, storage is a large cost in the cloud, with persistent storage volumes being a large cost line in any cloud budget. Storage housekeeping is a necessary evil, meaning you should be terminating, deleting, archiving or tiering unused or stale data and snapshots. Avoiding this can lead to large bills for storage you no longer need or use, particularly if you heavily utilise SSD volumes, these costing more than double their HDD counterparts.

6. Failing to innovate - In our opinion, one of the greatest benefits of public cloud is the wealth of innovation the providers and 3rd party providers bring. However, many users fail to capitalise on the benefits of services due to the sheer volume of services out there. For example, what if I told you I could reduce your storage bill in AWS by 20% and at the same time give you 20x performance? Using containers? You could potentially reduce your spend by 75%. Work with partners who do this for a living and use their innovation to drive down cost and complexity.

7. It just keeps growing… - Governance and visibility are two growing issues. Knowing who you are spending with, and who is consuming resource (often outside the controls of procurement and finance) are huge problems. We have seen bills grow from £10k per month to over £80k within a 6 month period, and finance don’t know until it is too late. 

8. Data Transfer & Bandwidth - The cost of transferring data into your cloud is usually free. But transferring it out is another matter altogether. Before deploying or migrating a service it can help to design it to minimise data transfer between cloud or other services and don’t forget to consider how your users and applications will be interacting with other SaaS platforms. If you’re not careful, data transfer can easily become a significant cost burden

9. Going it alone - This might sound like a bad sales pitch, but don’t go it alone, there are plenty of partners like us who can help you on your journey. We might not know everything, but a good partner can accelerate your success and help prevent you making mistakes that can cost you a significant portion of your cloud budget.

10. Simplify cloud cost management - Are you using massive spreadsheets to track and manage cloud spend? If yes, please stop, save your team this pain and automate the process with a tool that lets you allocate, track cloud spend, creates automated budget reviews and more. Doing this will give you intelligent analytics that will transform your understanding of your cloud spend.

In conclusion, controlling and optimising your cloud spend doesn't have to be difficult. A lot of the success comes down to teamwork, processes and culture as well as using the right tools for your particular situation.

If you'd like some help with this, feel free to get in touch :-)