The hot topic of the week is generative AI. With the release of Microsoft Copilot, we should finally start to get some real-world information and data on questions the industry has been asking for almost a year: How useful is gen AI, really? How transformative will it be, and how quickly? Is it worth the cost?
I’m not going to predict the answers to those questions, but this Wall Street Journal article, and this No Jitter interview with Tim Banting of Omdia offer some perspectives, especially around the question of ROI. Microsoft Copilot costs $30 per user per month, as does Google’s gen AI-powered assistant, called Assistant. In contrast, Zoom’s AI Companion comes at no additional charge, making the ROI case moot for Zoom users. The WSJ quotes a Forrester report that noted a savings of two hours a month recoups the $30 cost for an employee making as little as $15 an hour.
The ROI question gets more serious as deployments scale up. Microsoft sells Copilot with a minimum 300 licenses; at $30 a month that’s $9,000, which indeed seems reasonable. But across the whole enterprise—say 10,000 knowledge workers--at $30 a month that’s $300,000, or $3.6 million a year, which is real money even for a very large enterprise to spend for any new tool, let alone a technology whose payback is still relatively unproven.
So then the question becomes, who gets the cool new AI feature first?
A large enterprise with a vast knowledge worker base will almost certainly plan to phase in any gen AI assistant rollout and use the pilot period to understand the impact of these tools on its specific types of workers. One thing I hope we see over the next six months is some baselining of what sorts of jobs benefit the most from these AI-driven assistants. You might naturally think of sales since it’s revenue-generating, and marketing since it relies on creative processes that, anecdotally, gen AI seems capable of shortcutting.
I wonder how contentious this might get, this question of who gets gen AI assistants and who doesn’t. Essentially you’re telling someone: Unlike your colleague over there, your time is not valuable enough to spend $30 a month on.
Normally you wouldn’t expect people’s noses to get out of joint because they didn’t get the latest IT tool. But thanks to ChatGPT, AI has captured the imagination of people throughout society and, therefore, throughout the workforce. It’s not the same as being denied the latest Windows upgrade. It’s more like if, back in 2007, your company had bought the first iPhones for half the people, and not for the other half.
And it’s not just a matter of discontent. Denying people the technology they crave leads to Shadow IT. If people see Generative AI saving their colleagues time, they’re going to get it for themselves, one way or another.
So as gen AI-driven assistants begin to roll out within enterprises, IT can expect the usual range of headaches with new technology deployments, likely amplified and accelerated by the extreme hype we’ve seen over the past year when it comes to AI.
One way for you to get a handle on these challenges is to attend Enterprise Connect 2024 next March. I’m excited about how our program will be covering the impact of AI-driven applications: Our AI & Automationtrack will feature three sessions highlighting enterprise end user speakers discussing their AI strategies; and Kevin Kieller of EnableUC will present adeep diveon Microsoft Teams and Copilot. And of course we’ll be addressing AI’s use for the contact center throughout our CX track, including a case study session with leaders from Medtronic.
Those enterprise end user-led sessions will be among more than a dozen such sessions led by your enterprise peers. The EC 2024 program features over 20 speakers from large enterprises including Delta Air Lines, Weyerhaeuser, U.S. Bank, EY, Yum! Brands, Cushman & Wakefield, and many more. It’s an amazing lineup of truly awesome professionals, and it’s your best chance all year to learn from and network with your enterprise peers. I hope you canjoin us in Orlando!
