Analysis Although a lot is promised of generative AI, it has the potential to be expensive at scale, and the return on investment isn’t always clear. It’s understandable why some enterprises, big and small, may be hesitant to invest in the technology at this stage.
However, according to Google — a mega-corp whose generative AI investments depend on users buying into the tech — that’s exactly what enterprises should be doing: Buying into the tech.
“Our research shows that early adopters of gen AI are reaping significant rewards from increased revenue, to better customer service, to improved productivity,” Oliver Parker, VP of global gen AI go-to-market at Google Cloud, said in a canned statement.
The findings are part of a National Research Group (NRG) report commissioned by Google, which found that early adopters of generative AI were already seeing boosts to revenue. How much more revenue? Well, of the 2,508 C-suite executives surveyed, 61 percent (1,533) said they’d already deployed gen-AI tools in production, and, of those, 86 percent (1,318) reported estimated revenue gains of more than six percent.
When we asked, Google told us that estimate is “based on past / existing gen-AI initiatives.”
The report highlights several areas where generative AI is having a tangible impact, including productivity, security, business growth, and user experience.
Of respondents, 1,097, or 43 percent, reported that gen AI had a meaningful impact to productivity. NRG found that roughly 45 percent of that subset — about 20 percent of total respondents — estimated that these tools had more than doubled employee productivity.
The report also found that roughly 39 percent of respondents (975) had managed to grow their business using generative AI tools, and of them 77 percent (751) reported improved leads and new customer acquisition from the use of the technology.
It was a similar story with regard to user experience, where of approximately 37 percent of respondents (921) indicated that generative AI tools had had a meaningful impact. Of that group, NRG reported that 85 percent (782) saw increased engagement in the form of higher traffic, click through rates, or time on site, while 80 percent (737) reported higher customer satisfaction.
Finally, with regard to security, generative AI showed promise for sussing out and resolving security incidents. According to NRG, 56 percent (1,405) of executives reported that gen AI had in some way bolstered their org’s security posture, and 82 percent of that subset cited improved threat detection, and 71 percent reporting reduced time to resolution.
We’ll note that many of these figures didn’t appear in the original report and were provided by Google after The Register expressed concern that many of the claims were presented as a percent of an unknown quantity of respondents.
As it stands, the survey found there remains a considerable gap in generative AI adoption, with 39 percent of enterprises having not yet implemented the technology in production.
If you reckon that figure seems low, you’re not alone. As we previously reported, as of February, just 5.4 percent of US businesses reported using AI, according to the Census Bureau. Even if you just look at the information sector where adoption reached 18 percent, that’s still far less than the 61 percent cited in Google’s NRG report.
What is clear from the NRG survey is that 47 percent of respondents now plan to use generative AI to develop new products and services, and 49 percent intend to leverage any gains heralded by the tech to bolster profit margins.
The search for AI ROI
Google is far from the only corp trying to convince customers that this whole AI thing is going to be worth it in the end. All of the major cloud providers, including Microsoft and AWS, have invested massive sums of capital in GPUs and other accelerators necessary to train and/or facilitate the training of AI models and services, let alone inference.
And to realize a return on their own investments, they need to convince folks that generative AI is going to save them money, boost their profits, or even just cut headcount, as IBM CEO Arvind Kirshna previously suggested when he said the quiet part out loud. So it’s no surprise that these companies are keen to highlight any and every AI proof-of-concept that shows even a modicum of potential.
For instance, when speaking to analysts on Amazon’s Q2 earnings call, CEO Andy Jassy claimed that its generative AI code assistant dubbed “Q” had saved the e-commerce giant the equivalent of $260 million.
“With Q’s code transformation capabilities, Amazon has migrated over 30,000 Java JDK applications in a few months, saving the company $260 million and 4,500 developer years compared to what it would have otherwise cost,” he boasted.
As impressive as that might sound, it’s worth remembering that, in its 2023 fiscal year, Amazon made $30.4 billion in profit. Savings of $260 million equate to less than one percent of that.
Meanwhile, Microsoft has been among the most aggressive at building generative AI features into its various products, often at an added expense. GitHub Copilot and Copilot for Office 365 are the two most obvious examples of Microsoft’s efforts to commercialize its software and infrastructure investments.
But, at least as of March, prospective customers remain unconvinced the extra $30 per user is worth it. This is despite another commissioned report claiming that Copilot testers worked 29 percent faster, and 77 percent who used it for two weeks couldn’t quit it.
But, while AI service providers talk up proofs of concept and ROIs, the folks charged with advising enterprises on AI are sending a different message. Chatting to The Register last month, Gartner analyst Frances Karamouzis suggested that CFOs not even bother with calculating AI ROI because it’s not something that’s necessarily going to be demonstrable on their financials. ®