Cloud pricing hits new mark amid multiple drivers

Cloud pricing on the rise

Recent moves from cloud providers underscore the trend toward rising prices. IBM Cloud earlier this month announced IaaS and PaaS price hikes that will go into effect Jan. 1, 2024. IaaS customers outside the U.S. will see their location premiums increase, with Brazil topping the list with a 7.5% uptick. All customers will see a 3% increase in PaaS pricing, as well as a 25% uplift for Accelerated Archive, an optional feature for IBM Cloud Object Storage.

Other signs of rising cloud costs come from Salesforce and ServiceNow, among other providers.

Customers can expect such price increases to continue, as the public cloud remains a key element of organizations' IT modernization and business transformation strategy, said Sid Nag, vice president analyst at Gartner.

Public cloud providers will strive for price parity against the competition, and this is the time to do it, while the peers increase their pricing structure.

"Public cloud providers will strive for price parity against the competition, and this is the time to do it, while the peers increase their pricing structure," Nag said. "Moreover, the addition of GenAI will provide the fuel, the impetus and justification to do so."

GenAI's emerging role

Indeed, the emergence of generative AI is influencing cloud pricing -- along with more traditional variables such as IT salaries and data center energy costs.

Salesforce, for example, cited the addition of features including "generative AI innovations" when announcing its price increase, which the company said was its first in seven years. ServiceNow, meanwhile, plans to introduce generative AI capabilities in premium versions of its IT Service Management, Customer Service Management and HR Service Delivery products, which the company said it will price higher than the Pro versions.

The expenses of running large language models (LLMs) and cooling massive DPU and GPU clusters rank among the costs cloud providers might pass along to customers.

But some cloud providers might not be eager to do so.

Vikrant Karnik, cloud and technology services leader at Genpact, said the cloud companies he's spoken with plan to initially give away generative AI services as part of their broader offerings -- bundled together under a flat fee. Genpact is a professional services firm based in New York City.

"They will give you the access as much as possible, because they know that people are trying it out," he said.

But that approach will change once cloud providers start to see higher use of their generative AI services, Karnik noted. At that point, they will begin to charge customers a premium for attributes such as response time. For example, providers will ask customers to pay more to prioritize their LLM queries ahead of other customers, he said.

But as premium pricing for generative AI becomes more commonplace, customers will need to consider whether a given service is worth the increased cost.

"The added costs for embedded AI in products can be based on cloud or on-prem infrastructure costs, but it can also be part of the hype cycle, which has seen the valuations of AI-based companies being much higher," said Bret Greenstein, data, analytics and AI leader at consultancy PwC.

The important consideration is determining the value AI provides when embedded in a particular application, Greenstein said. If the AI component only provides a natural language interface to an existing app, it isn't worth a high premium, he said. On the other hand, an AI feature could prove its value if it transforms how an organization works and, as a result, generates significant labor savings.

Other cloud cost factors

Higher wages, rising energy demands and generative AI aren't the only cloud cost drivers. They're not even the most important ones, according to some industry executives.

Shane Quinlan, vice president of product at Kion, a Columbia, Md., company that provides financial management and other cloud tools, works with U.S. government agencies. He cited that sector's IT spending shift to the cloud as the root cause of increasing costs.

"Will hyperscalers raise prices? Sure. Is that the primary driver of cost for federal agencies? No. The primary cost driver is usage," Quinlan said.

In Quinlan's view, cloud costs are rising because more agencies are devoting more of their budgets to take advantage of the technology.

Cloud consumption, however, isn't always managed with care.

"The No. 1 thing that drives cost in cloud is waste," said John Purcell, chief product officer at DoiT International, a cloud cost management technology and services provider based in Santa Clara, Calif. The company offers an anomaly detection product designed to flag unexpected departures from a customer's typical cloud spending pattern.

"There's an expression that says your costs are not driven by the workloads you run in the cloud -- they're driven by the infrastructure you forget to turn off," Purcell noted.

Purcell said the higher-cost-through-poor-management driver has been around since the advent of the cloud, but it became more of an issue when adoption accelerated in the early 2010s. More recently, COVID-19 sparked another wave of cloud uptake as organizations quickly embarked on digital transformation programs.

"It was just a lot of overreach, [and] costs got out of control, not surprisingly," Purcell said.

Cloud inflation is indeed a real phenomenon, he said. But the current focus on cloud costs is just the latest in a series of recurring economic tests and cost struggles.

"These operating challenges just continue," Purcell said. "They're cyclical."

John Moore is a writer for TechTarget Editorial covering the CIO role, economic trends and the IT services industry.