According to the latest data from analyst Gartner, the data center market is surging by 10%, driven primarily by hyperscalers deploying artificial intelligence (AI) servers.
AI in enterprise systems is touted as a way to increase productivity and improve efficiency, but running AI workloads is expensive, and someone has to pay for it.
Given that the average cost of an AI server is $32,000, noted Gartner analyst John-David Lovelock points out that a rack of AI servers can cost more than $1 million. While hyperscalers have the capital to build large AI infrastructures and competition among hyperscalers will keep prices down for AI services, the same will be true for software-as-a-service (SaaS) providers. Not applicable.
When Computer Weekly raised the issue of AI price increases in a recent conversation with Salesforce, Jayesh Govindarajan, senior vice president of AI, said: Go and do it. If he truly provides value and is making someone 10x more efficient, people will be willing to pay for it. ”
Brian Chess, senior vice president of AI at NetSuite, said at a recent SuiteConnect London event that the company benefits from partnerships with Nvidia and Oracle, and that the graphics needed to run AI workloads on Oracle Said to provide access to the processor unit (GPU). cloud. However, he acknowledged the costs associated with operating an AI infrastructure and said Netsuite would need to evaluate whether to charge additional fees to justify further investment in AI. “We are looking at price and business advantages,” Chess said. “We're seeing some really great marriages that are worth the cost.”
AI is starting to look like an arms race in enterprise software. Lovelock said companies are starting to offer AI capabilities for free, which means it becomes a fixed cost that impacts the bottom line. “Enterprise software companies using the cloud have no choice but to incur this cost, and providing the compute power to run AI means higher operational costs,” he said. Stated.
Lovelock said the challenge for SaaS providers is that this cost fluctuates. “We don't yet know what the long-term average usage per user will be. So right now, users are in a bit of a volatile situation,” he said. “They either charge a reasonable amount for his AI capabilities or they deduct it from their profits. Either way, the cost as a percentage of revenue is fixed.”
Gartner has identified a number of categories in the enterprise software market where true competition exists. According to Lovelock, this means SaaS providers will have to offer his AI for free or as an add-on that IT decision makers can choose to purchase.
“Costs can fluctuate widely,” he said, adding that over the next year, SaaS providers will be busy evaluating average usage per person and the average cost to meet that usage. “They have some great models and some great estimates, but we're in uncharted territory,” Lovelock said.
Looking at a hypothetical example of how two competing SaaS providers might face off on AI capabilities, he said: “The first question is, can we change the price?” For example, if Microsoft significantly increases the price of its Outlook email service to include Copilot AI capabilities, Google's AI could be free or low-cost. There is a risk that IT buyers will push back and look to alternatives such as Gmail, Lovelock said.
“The market doesn't give them pricing power,” he says. “SaaS providers need to determine their corporate strategy.”
While SaaS providers that have traditionally focused on being low-cost providers in the market may not offer the latest features, Lovelock said they need to offer AI to survive, and such He said that other companies would not have the pricing power to provide that. It's about building AI into their core products and charging everyone more.
Instead, some providers may offer AI as an add-on. This will allow these software providers to offer AI capabilities to customers who are hungry and willing to pay for them, he said. “For people who don't need AI, there's no harm and there's no price increase,” Lovelock said. This creates a second layer of non-AI-enabled software.
For example, SAP has started offering customers a 50% discount to switch from on-premises enterprise resource planning, although it makes its AI capabilities available only on its Rise cloud platform.
Justification of additional costs
“Some of our modules have more AI, and some have less,” NetSuite's Chess said.
While those with higher levels of AI functionality come with a price tag, we suggest that AI-enabled modules use new functionality to provide tangible benefits that justify the additional expense customers expect. did.
For example, Netsuite Analytics Warehouse provides AI-powered analytics that uses machine learning to uncover optimization opportunities. Netsuite EPM (Enterprise Performance Management) is now available in Europe, the Middle East, and Africa with AI-powered predictive modeling, which the company says allows customers to run “what-if” models.
Lovelock said that in a stable market with strong competition, SaaS providers will focus on customer retention. Providing AI capabilities at your core can be a necessary cost to compete effectively with your competitors.
