This audio is auto-generated. Please let us know if you have feedback.
Enterprises are grappling with escalating software spending as vendors add generative and agentic AI features to their products and adoption efforts advance. The technology is driving up costs across infrastructure, data management and applications, according to West Monroe.
More than 9 in 10 executives expect AI adoption to increase technology budgets in the coming year, with software as a major culprit, the consulting firm found in a report published Monday. West Monroe surveyed 310 executives in procurement, IT and finance at organizations with at least $100 million in annual revenue.
Nearly half of organizations saw licensing and subscription costs for renewals and new contracts increase by more than the industry average of 10%, the survey found.
As costs rise, so does the pressure on CIOs to demonstrate a return on software investments, Dhaval Moogimane, leader of West Monroe’s high tech and software practice, told CIO Dive. “The real challenge is tying spend back to value creation,” he said. “Executives often do not have visibility into how much others are spending, so they are working off of limited information.”
Global spending on AI is snowballing, driven by huge hyperscaler infrastructure investments and the proliferation of software tools powered by the technology. Enterprise customers are on the hook as costs flow downstream to end users.
“AI software is expensive to develop and maintain due to high computational needs, specialized talent and continuous upkeep, leading some tech companies to raise prices to offset these costs and compress profit margins,” Jonathan Selby, tech industry lead at risk management firm Founder Shield, said in an email. “However, this trend is nuanced, as competition and open-source models may eventually make some AI services more accessible.”
The massive wave of vendor investments has yet to crest.
Last week, Microsoft pledged $30 billion to build out its Azure cloud empire in the U.K. and added $4 billion to an existing $3.3 billion investment in Wisconsin data center facilities. In July, Google Cloud tacked $10 billion onto its plan to pour $75 billion into capacity expansions this year. Meanwhile, cloud behemoth AWS committed $10 billion to North Carolina AI data center buildouts in June, as the largest hyperscaler plans to invest $100 billion this year in infrastructure-related capital expenditures.
Cloud providers are footing the bill for a lion’s share of AI investments, which will push the global market by nearly 50% year over year to almost $1.5 trillion in 2025, according to Gartner projections. But the mix is shifting toward enterprises and the software they consume.
Inundated with AI
Enterprise spending can be hard to track, as AI spreads throughout IT ecosystems, according to Erik Peterson, founder and CTO at IT cost management platform provider CloudZero.
“AI spend is coming in from so many different directions,” Peterson told CIO Dive. “It’s coming in in the form of your own software engineering or in-house teams who are using AI to accelerate their work, and it’s replacing functionality in existing systems.”
Poor visibility into costs can cloud perceptions, influencing critical investment decisions. Nearly two-thirds of executives surveyed by West Monroe said they pay more for software assets compared with their peers: only 3% believed their bills were lower.
Organizations that think they are overspending adopt defensive rather than strategic procurement strategies, prioritizing cost-cutting over other business objectives, according to West Monroe.
Some of the spending excess is tied to vendor contract negotiation missteps. Many companies are swallowing double-digit price bumps for software renewals despite rate increases of less than 10% for Oracle on-premises, SAP support and Salesforce renewals, according to analytics firm Green Cabbage research cited by West Monroe.
Two major enterprise vendors have already signaled AI-related cost changes. Salesforce bumped up prices across several of its enterprise offerings by an average of 6% beginning last month. Earlier in the year, the company raised monthly per-user costs for Slack business plans by 20% and instituted new pricing for its Agentforce platform.
Microsoft axed volume licensing program discounts under Enterprise Agreement and Microsoft Products and Services Agreement licenses for many of its cloud-based software suites, including Microsoft 365 and Dynamics 365, in August.
Overprovisioning and shadow IT is also inflating software budgets, Moogimane said.
“A lot of organizations made significant software investments over the past few years — particularly with cloud agreements — based on certain expectations of business growth or agility,” said Moogimane. “Some companies also invested in higher levels of resiliency or support agreements than they currently require.”
As long as AI adoption remains in its early stages, IT executives will have to juggle multiple priorities, Moogimane added.
“Business units are experimenting with AI tools on their own, often outside of central oversight,” said Moogimane. “You don’t want to discourage that type of experimentation because it is where a lot of innovation and ground-up adoption comes from, but it will create a proliferation of software.”
