Connect with our experts
Share a few details about your business challenge, and we’ll get right back to you.
Connect with our experts
Share a few details about your business challenge, and we’ll get right back to you.
Agentic AI continues to dominate headlines and boardroom conversations, promising transformative productivity gains for organizations worldwide.
SoftwareOne Co-CEO Melissa Mulholland was on Bloomberg News this week, talking about how AI is undeniably transformative, but many companies underestimate the implied costs and complexities of implementation.
SoftwareOne is the enabler of AI, advising clients on how to best deploy solutions to reap productivity gains.
Here are a few excerpts from the conversation with The Opening Trade anchors Anna Edwards and Guy Johnson.
Anna Edwards: We are looking for clues, always as to how AI is actually making a difference on the ground for corporates. … Will it deliver on productivity?
Melissa Mulholland: AI is transformative, that's without a doubt. However, most companies don't think about the implied costs. It takes discipline. It takes decisive action as CIOs across the world are faced with multiple cost additions.
For example, when you have agent sprawl, so multiple agents being deployed across an organization, data governance, security, also cloud costs. So oftentimes it's not just about implementing an agent, and it solves and delivers return on investment, but in fact, oftentimes companies experience higher costs because of the other aspects that they have to think about with people.
Guy Johnson: There's this whole expectation at the moment that companies are going to deliver massive productivity as a result of agentic AI, the use of AI. You're telling me that actually looks harder to achieve than maybe the … zealots of the AI world are suggesting?
Mulholland: Absolutely AI will deliver productivity gains and we know that to be a fact. But it also requires holistic business planning. Oftentimes, organizations don't think about that end-to-end because you have to think about employing your people, you have to make sure that those people know how to use the agents, as well as to maintain the governance associated to it. So you have to look at the wholistic cost and oftentimes companies just assume implementing AI they're going to have return on investment immediately.
Johnson: … You're saying it's more expensive at the moment to use this stuff than maybe to employ humans? Am I correct in that assumption?
Mulholland: It depends on the use case and it depends on the model. So if you have a very specific use case you've planned out, and you've organized your people to be able to deploy it, then it can deliver a return on investment.
Johnson: How many companies have you seen doing that well?
Mulholland: I would say probably 30% if I were to articulate. And most companies don't think about the wholistic aspects. So for example, we have a Swedish finance company where they actually deployed agents. They resulted in 13% revenue savings, but they overspent their budget by 80%.
So that's what can happen if you don't have an articulated plan and businesses aren't thinking about how to ensure that their people are trained and are using the AI agents.
Edwards: So what are these quiet costs then, Melissa, that get into this process that are not foreseen?
Mulholland: Security, data security, data governance, also making sure all of your data is in the cloud. There are reasons that hyperscalers and companies want to have these agents be deployed. So making sure that your data is in place is key. Also retraining and ensuring that people, the employees, know how to use the agents is vital because at the end of the day, it's a change, it's adoption and change management. For all of us, learning how to prompt and use this to get ROI is essential, and that is a key component of it.
Click here to watch the full interview.
Author
