BOZEMAN, Mont. – April 15, 2025 – AI knowledge cloud firm Snowflake (NYSE: SNOW), in collaboration with Enterprise Technique Group, at this time launched the “Radical ROI of Generative AI,” a analysis report surveying 1,900 enterprise and IT leaders throughout 9 totally different nations — all of whom are actively utilizing AI for a number of use instances.
Of all respondents, 92% reported that their AI investments are paying for themselves, and 98% plan to take a position extra on AI in 2025. As AI adoption accelerates throughout world enterprises, a strong knowledge basis has emerged because the cornerstone of profitable implementation, but respondents are nonetheless grappling with tips on how to make their knowledge AI-ready.
Researchers from Enterprise Technique Group recognized, and carried out deeper analysis between Nov. 21, 2024, to Jan. 10, 2025, with early adopter organizations — these already augmenting and executing enterprise processes in manufacturing, utilizing industrial and open-source fashions somewhat than consumer-grade, subscription software program comparable to ChatGPT. Of three,324 respondents, 1,900 (57%) mentioned they’re utilizing industrial or open supply generative AI options. Extra particulars round methodology will be discovered inside the report.
“I’ve spent virtually twenty years of my profession growing AI, and we’ve lastly reached the tipping level the place AI is creating actual, tangible worth for enterprises throughout the globe,” mentioned Baris Gultekin, Head of AI, Snowflake. “With over 4,000 prospects utilizing Snowflake for AI and ML on a weekly foundation, I routinely see the outsized influence these instruments have in driving larger effectivity and productiveness for groups, and democratizing knowledge insights throughout complete organizations.”
Early AI investments are proving to achieve success for almost all of enterprises, with 93% indicating that their AI initiatives have been very or largely profitable. In reality, two-thirds of respondents are beginning to quantify their generative AI ROI at this time, discovering that for each greenback spent, they’re seeing $1.41 in returns (a41% ROI) via price financial savings and elevated income.
Nonetheless, there are world nuances round the place organizations are focusing their AI efforts that instantly correlate to every nation’s AI maturity, and their outcomes when it comes to driving ROI throughout areas:
- Australia and New Zealand (ANZ) respondents have seen a 44% return on their AI investments. In comparison with the worldwide common, organizations in ANZ have been extra more likely to cite enhancing buyer satisfaction as a key objective for his or her AI initiatives (53% versus 43%), and fewer more likely to prioritize internal-facing initiatives (47% versus 55%).
- Canada respondents have seen a 43% return on their AI investments. Canadian organizations have been extra more likely to say that they’re solely pursuing preliminary AI use instances (45% versus 36%), suggesting that many are earlier of their AI adoption journeys than world counterparts.
- France respondents have seen a 31% return on their AI investments. In comparison with the worldwide common, French firms are much less more likely to practice or increase massive language fashions (LLMs) with proprietary knowledge utilizing retrieval-augmented technology (RAG) (59% versus 71%), suggesting a lag in maturity for his or her AI methods.
- Germany respondents have seen a 34% return on their AI investments. German organizations have been extra more likely to report challenges with infrastructure, notably in assembly storage and compute necessities for AI (69% versus 54%).
- Japan respondents have seen a 30% return on their AI investments. Japanese organizations differed of their strategic objectives for AI, being least more likely to focus their AI efforts on customer support and assist (30% versus 43%) and monetary efficiency (18% versus 30%), however the most probably to harness AI to assist reduce prices (43% versus 32%).
- South Korea respondents have seen a 41% return on their AI investments. South Korean companies are using mature AI use instances, reporting the best use of open supply fashions (79% versus 65%), and usually tend to practice or increase fashions with RAG (82% versus 71%).
- United Kingdom respondents have seen a 42% return on their AI investments. By way of strategic objectives, UK-based organizations have been extra more likely to prioritize the worth AI brings to finish customers, with respondents beating the worldwide common in citing each operational effectivity (57% versus 51%) and innovation (46% versus 40%) as main enterprise drivers.
- United States respondents have seen a 43% return on their AI investments. American firms led in profitable AI operationalization, with respondents extra typically than every other nation to say that they’ve been “very profitable” at operationalizing AI to attain their enterprise objectives (52% versus 40%).
Many organizations additionally report that they’re grappling with troublesome choices to construct on the momentum. Respondents reported challenges with figuring out probably the most impactful use instances and elevated stress to make the proper choices — all whereas grappling with restricted sources:
- Too many use instances, too few sources: 71% of early adopters agree they’ve extra potential use instances that they wish to pursue than they will presumably fund.
- Resolution-making blind spots: 54% agree that deciding on the proper use instances primarily based on goal measures like price, enterprise influence, and the group’s potential to execute is tough.
- Aggressive stress mounts: 71% acknowledge that deciding on the mistaken use instances will harm their firm’s market place.
- Job safety considerations come up: 59% of respondents say that advocating for the mistaken use instances might price them their job.
Organizations are more and more incorporating their proprietary knowledge to maximise AI’s effectiveness, with 80% of respondents selecting to fine-tune fashions with their very own knowledge. Regardless of this widespread recognition of information’s significance — with 71% of respondents acknowledging that efficient mannequin coaching and fine-tuning requires multi-terabytes of information — important challenges persist in making this knowledge AI-ready. With the bulk struggling to utilize their most beneficial asset, organizations declare that the next are the most important knowledge hurdles for driving AI success:
- Breaking down knowledge silos: 64% of early adopters say integrating knowledge throughout sources is difficult at this time.
- Integrating governance guardrails: 59% say implementing knowledge governance is troublesome.
- Measuring and monitoring knowledge high quality: 59% say measuring and monitoring knowledge high quality is troublesome.
- Integrating knowledge prep: 58% say making knowledge AI-ready is a problem.
- Effectively scaling storage and compute: 54% say it’s troublesome to fulfill storage capability and computing energy necessities.
There’s a important alternative for companies to beat these challenges and unlock the complete potential of their knowledge for extra correct, related, and impactful AI outcomes with a unified knowledge platform.
“The fast tempo of AI is barely accelerating the necessity for organizations to consolidate all of their knowledge in a well-governed vogue,” mentioned Artin Avanes, Head of Core Knowledge Platform, Snowflake. “Having a simple, related, and trusted knowledge platform like Snowflake is crucial not only for serving to customers see sooner returns on their knowledge investments, nevertheless it lays the muse for customers to simply scale their AI apps in a compliant and safe method — with out requiring specialised or exhausting to search out technical expertise. A managed, interoperable knowledge platform gives seamless enterprise continuity as world enterprises faucet into their complete knowledge property to steer within the evolving AI panorama.”
The total analysis report is right here: “Radical ROI of Generative AI.”
