How AI Can Help Your Company Get a Handle on Cloud Costs
As AI costs balloon, companies are feeling the pinch.
The average business has seen its cloud spending jump 30% in the last year, with generative AI the big culprit. Horror stories aren’t hard to find, like racking up a $30,000 bill in 12 hours due to calls to AI services. Frontline practitioners, meanwhile, are overwhelmed, with more than 70% saying GenAI-driven cloud spending has become “unmanageable.”
It’s the latest twist on a stubbornly persistent problem: unpredictable and out-of-control cloud costs. Spending on public cloud services is set to top $800 billion in 2024, a 20% jump from the year before. Worse, more than three-quarters of businesses say that anywhere from 20%-50% of their cloud spend is wasted.
It doesn’t have to be that way. AI may be driving up costs, but it also holds the key to controlling them for companies grappling with soaring cloud budgets. Here’s why cloud costs are so high—and how businesses can bring them down to earth.
Why Cloud Costs Are Through The Roof
How did we get here? The cloud’s allure is also its Achilles heel: Providers make it very easy for anyone in the company to spin up new resources with just a few clicks. But keeping track of costs is another story.
Plus, even if a company does have a clear picture of their cloud spending, reining it in isn’t easy. Most businesses resort to a sort of whack-a-mole operation—manually identifying and addressing excess spend in a never-ending game of catch-up.
Then there’s the sheer complexity of cloud offerings, which can make cell phone plans seem simple by comparison. A major provider like AWS has hundreds of different products and services, each with hundreds of possible variations. Identifying the most cost-effective option becomes a full-time job.
Even those of us who should know better can lose track of cloud spend. At my first software company, we built a new product but didn’t factor in the cloud costs of running it, then had to rewrite the whole product with a new design.
The latest AI applications, with their heavy processing load, only add fuel to that fire. Expensive and unnecessary calls to AI services can easily swell into the hundreds of thousands of dollars.
FinOps: A New Way Of Looking At Cloud Costs
The way forward? At a high level, getting a handle on cloud costs starts with bringing together two teams that are usually miles apart: finance and engineering.
This is where the emerging discipline of FinOps (financial operations) comes into play. Smart CFOs, frustrated with spiraling spending, are building bridges with developers on the front line. By giving developers more visibility into cloud spending and holding them accountable, companies are seeing dramatic improvements in savings and efficiency.
But to make that happen, something crucial is often missing: the right tools. New technologies, including AI, are finally giving companies the power to fight back against surging cloud costs. Here are three critical steps to take to get cloud costs under control with new technology:
Step 1: Better Understand Your Spending
Cost management tools for the cloud aren’t new. But until recently, most were complex to configure and demanded constant maintenance. At best, they flagged overruns, leaving it to teams to address them (or not). But AI and automation have made identifying sources of wasted spending dramatically easier.
The most innovative cloud cost management tools let anyone ask questions in plain language and get an equally clear answer. What will the cloud costs for a new product feature be? Which teams are consuming the most cloud services? What are the top three steps the development team can take to reduce cloud spending?
Calculating this in the past required diving into a dozen spreadsheets, but this is what AI was built for.
Step 2: Identify What Actions To Take
Having visibility into cloud spend is one thing. Knowing what actions to take to optimize it is an equally critical step. New tools enable developers to not only identify overruns, but also get recommendations on the right fixes based on past patterns and resolutions. That might mean turning off idle resources, or optimizing storage costs by moving seldom-accessed data somewhere cheaper.
It turns out, however, that having a clear path forward isn’t always enough. In fact, the latest FinOps Foundation survey shows that while automation is a priority for companies, the vast majority still have people doing fixes manually. On a busy software team, who’s got time for that?
Step 3: The Missing Link—Automate Cloud Fixes
The best cloud cost management tech now goes one step further—actually automating those fixes and reducing spend automatically.
Intelligent auto-stopping is one relatively simple example. By setting cloud resources to stop running when not in use—much like a motion-sensing light switch that turns on and off when someone enters or leaves a room—engineering teams no longer have to worry about wasting money if someone forgets to shut them down.
New AI-powered tools can also forecast spend, automating purchases of the most economical plans from cloud providers to match spending patterns. Rather than a harried team lead having to comparison shop for the best AWS plan, new tech anticipates what’s needed and secures the lowest price.
Governance is another area where AI is proving a difference maker. To help optimize costs, AI-powered tools let developers create and enforce rules for their cloud environments, ensuring entire teams abide by the same standards and no one goes rogue.
For businesses, the savings from AI and automation can be substantial. I’ve seen Fortune 1000 companies shrink cloud spending by as much as 50%.
The reality is that the cloud costs challenge affects businesses big and small—and AI has only raised the stakes higher. The good news: By harnessing AI and automation to keep a lid on expenses and avoid nasty surprises, companies can breathe easier when opening that cloud bill.