Snowflake Report Reveals Surging ROI from Generative AI Adoption

Snowflake, the AI Data Cloud company, in collaboration with Enterprise Strategy Group, today released the “Radical ROI of Generative AI” report, which revealed that 92% of respondents stated their AI investments are already generating returns, with 98% planning to increase their AI spending in 2025.

The research surveyed 1,900 business and IT leaders across nine countries, including the UK, France and Germany — all of whom are actively utilising generative AI for one or more use cases.

Just two-and-a-half years since generative AI began dominating tech discussions, early adopters are achieving tangible success with both internal and external applications. Over half (55%) of respondents prioritised employee-facing solutions to boost productivity and efficiency, while 44% initiated customer-facing implementations to enhance customer experience and satisfaction.

“I’ve spent nearly two decades developing AI, and we’ve now reached a tipping point where it’s delivering real, measurable value for enterprises globally,” stated Baris Gultekin, Head of AI at Snowflake. “With over 4,000 customers using Snowflake for AI and ML weekly, I regularly witness the significant impact these tools have on improving team efficiency and democratising data insights across entire organisations.”

Mohamed Zouari, General Manager for the Middle East, Africa, and Turkey at Snowflake, commented: “With the UAE projected to gain US$96 billion from AI by 2030 and Saudi Arabia launching a US$100 billion AI initiative, AI is fast becoming the blueprint for business growth in the Middle East. But without a data strategy, there is no AI strategy. Regional businesses face real challenges — from fragmented data infrastructures to talent shortages. At Snowflake, we’re supporting organisations to build the secure and scalable data foundations necessary to truly unlock AI’s potential.”

Driving Success Through AI Investments
The report reveals early AI investments are paying off for most enterprises, with 93% describing their generative AI initiatives as very or mostly successful. Respondents reported measurable improvements in efficiency (88%), customer experience (84%), and accelerated innovation (84%). In fact, two-thirds of those surveyed are already quantifying ROI, finding that for every US$1 million invested, they are achieving US$1.41 million in returns through cost savings and revenue gains.

As organisations progress on their AI journeys, they are allocating additional resources across key areas, including data (81%), large language models (78%), supporting software (83%), infrastructure (82%), and talent (76%). This reflects a broader strategic shift in how businesses prioritise innovation and future competitiveness.

Overcoming Data Barriers to Maximise AI Effectiveness
Realising AI’s full potential requires a strong data foundation. A growing number of organisations are turning to their proprietary data to enhance AI effectiveness, with 80% opting to fine-tune models using their own data. Despite this, many still face major hurdles in making data AI-ready, with the following barriers cited most frequently:

Unexpected costs: 96% of early adopters reported that one or more components of their generative AI solutions have cost more than initially anticipated. Additionally, 78% said at least half of their use cases have been more expensive than expected to implement.

Data silos: 64% indicated difficulties integrating data across sources.

Unstructured data: While unstructured data accounts for an estimated 80–90% of all data, only 11% of early adopters said more than half of their unstructured data is currently usable for training and fine-tuning large language models.

Governance challenges: 59% cited difficulties enforcing data governance policies.

Data quality: 59% reported challenges in measuring and monitoring the quality of their data.

Data preparation: 58% found making data AI-ready to be a significant obstacle.

Storage and compute: 54% said it remains difficult to meet storage capacity and computing power demands effectively.

Methodology
Enterprise Strategy Group researchers conducted in-depth research between 21 November 2024 and 10 January 2025, focusing on early adopter organisations — those actively augmenting and executing business processes in production using commercial or open-source generative AI models. Of the 3,324 respondents surveyed, 1,900 (57%) confirmed using commercial or open-source generative AI solutions, rather than consumer-grade software such as ChatGPT.