Enterprise AI buyers increase spending but demand proof

The Hoffman Agency has published research on enterprise AI buying trends, finding that buyers are increasing investment while subjecting purchases to heavier scrutiny.

The survey covered 450 senior technology decision-makers in the UK, US, France and Germany responsible for buying technology, equipment, software or services.

AI-related spending featured strongly in enterprise technology budgets over the past year. The research found that 45% of organisations invested in data and AI infrastructure, 44% in generative AI, 40% in cloud infrastructure, 39% in AI-enabled solutions and 34% in agentic AI.

Spending intentions also remain high. Over the next 12 months, 36% of respondents said they plan to increase spending on generative AI, while 34% pointed to agentic AI, data and AI infrastructure, and cloud infrastructure. Another 33% expect to raise spending on AI-enabled solutions.

The findings suggest companies are approaching AI as a practical business tool rather than a speculative bet. Improving operational efficiency was cited by 32% of respondents as a driver of investment, followed by keeping pace with technology advances at 30%. Reducing costs and faster product development were each named by 25%.

Those priorities are shaping where organisations deploy the technology. Task automation led at 38%, followed by workflow optimisation at 35%, product development at 34%, risk analysis at 31% and customer support at 30%.

Tom Broughton, Vice President and Head of Enterprise AI at The Hoffman Agency, said buyers remain focused on measurable returns.

"Despite all the hype around a multitude of benefits, enterprise AI buyers are still heavily focused on productivity gains. While perhaps understandable in the current economic environment, the AI growth narrative has yet to land with many enterprises," said Tom Broughton, Vice President and Head of Enterprise AI at The Hoffman Agency.

Rising scrutiny

Even as spending rises, the purchasing process is becoming more difficult. Nearly two-thirds of respondents, or 64%, said AI-related technology purchases face more scrutiny than other technology investments.

The research points to a market crowded with similar claims and limited points of comparison. Some 78% of buyers said many AI companies make claims that are difficult to verify or compare, while 69% said many sound similar and are hard to differentiate. A further 61% said they are unsure which products are genuinely credible because of the volume of AI offerings coming to market.

That uncertainty appears to be slowing decisions inside organisations. When claims are hard to verify, 46% said the decision process takes longer. Another 37% said larger or more established suppliers are favoured, while 32% reported increased procurement or compliance checks, and the same proportion said more stakeholders become involved in evaluation. More than a quarter, or 27%, said purchases are delayed, while 16% said deals are abandoned altogether.

"Buying enterprise AI solutions has become a full-body contact sport. Vendors are flooding the zone with unsubstantiated and incomparable claims," said Broughton.

"The result? More scrutiny, more scepticism, and more friction that is dramatically slowing purchasing and adoption. Enterprise AI buyers are yearning for evidence and credible claims that can help them make decisions with confidence," he said.

Proof points

Buyers identified several factors that increase confidence in suppliers. The most commonly cited was transparent explanations of how products work, selected by 46% of respondents.

Customer case studies with measurable outcomes followed at 41%, while 40% pointed to proof of real-world deployment at scale. Demonstrated technical performance benchmarks were cited by 37%, and 36% said independent third-party validation helps build trust.

The findings also show that outside sources remain important in the early stages of vendor selection. More than half of organisations, or 57%, use industry analysts to identify and shortlist suppliers. Business technology media was cited by 25% as a source for discovering new vendors.

Large language models are also beginning to influence the process. The study found that 17% of buyers use them to discover new vendors or solutions, and the same share said such tools influence decisions between vendors.

Broughton said the volume of AI marketing has made external validation more important in the buying process.

"A frenzy of AI marketing messages has confounded enterprise buyers. They're increasingly frustrated by the 'black box' nature of products and a lack of evidence-based success stories. They're turning to third-party experts to help in their assessments and shorten their processes. These are important lessons for marketers within vendors if they are to outperform competitors and meet growing market expectations for growth and profitability," said Broughton.