According to a recent IBM survey, more than 80% of CSCOs say that incorporating AI into their supply chains is a top priority. Sought-after capabilities include real-time demand sensing and inventory management — not surprising considering that 80% of the same group of CEO’s identified demand volatility as a top challenge.
AI conversations are also picking up due to some positive results flowing back from the front lines: McKinsey reports that AI-enabled early adopters have been able “to improve logistics costs by 15 percent, inventory levels by 35 percent, and service levels by 65 percent, compared with slower-moving competitors.” These are the kinds of results that will certainly get attention!
Fire Where There’s Smoke?
We’re guessing that despite all of this good press, 80% moving on AI may seem a little high to you. Well, it feels more than a little optimistic to us, too. Turning to sources who live and breathe the supply chain space, we find Robert Bowman at Supply Chain Brain, estimating that as few as 12% of supply chains use AI for everyday management decisions.
So yes, there’s clearly momentum in terms of plans and in understanding of the business case — but there’s still a big gap to fill. This got us wondering about the current state of this gap: between the CSCOs who say it’s a priority; the promising results from early adopters; and the reality that only a few supply chains have implemented AI in a material way?
What’s The Holdup?
Like everything else in supply chain, the answer isn’t simple or straight forward, but we can see a few major influences at play, currently refracting and reinforcing one another:
- Challenging Environment: It seems obvious but that doesn’t mean the pressure has let up: right now, there are simply too many challenges to choose from in supply chain. Inflation and geo political turmoil are both unrelenting, and both occupy a priority position and urgency which makes them poor candidates for longer term strategic implementations like AI. As Forbes says, AI isn’t a “panacea” for supply chain. (Forbes also cites numbers more akin to Robert Bowman’s in terms of positive results.)
- Trail Blazing is Hard: At this stage, no generally accepted playbook has emerged for how and where to get started with AI. We’re still in the early days, and it’s not as simple as pointing to the next module in SAP (for example). If you remember playing with Legos before they made all those models, sets, and plans… you may be familiar with the challenge!
- Failure Risk is Still High: The rate of AI project failure is estimated to be around 70-80%, based on reports from Gartner and TechRepublic. Supply chain isn’t a place with a large margin of error for new projects. …Especially right now!
The Case For Optimism, and Action
Despite these risks, we still see the upside — and of course, so do all those CEO’s planning to move into this space. The simple fact is, operational results like those reported by McKinsey can’t be ignored. They give strong evidence of a better mousetrap out there, so, at the CEO and strategic-investment level, the hunt is on. The story we linked to from Forbes also provides numerous examples of how AI can be (and has been) “extremely helpful.” And despite the gloomy success rates reported by Gartner — Gartner also reports that there’s no sign of slowing down AI investments — again, because the success stories offer proof that the tech can and does work. Use cases are diverse and plentiful enough now to suggest that AI for Supply Chain can be (much) more than worth the effort.
So…how can leaders make sure they get the good of AI in supply chain without the bad?
Our strong belief is that successful efforts are usually the outcome of good planning. We put together a white paper: Five Keys for AI and Supply Chain Success, for just this purpose. The first of these keys is probably the primary root of most successes and failures, and that’s selection of the use case / problem to solve. Hesitation at this step is likely also a major contributor to that gap between intention and action we discussed earlier.
We’ll be continuing this series as a way of expanding on the discussion we started with the white paper. We’ll be talking a lot about data, and modern cloud technology. And yes, we’ll present the case for AI innovation. We believe that modern cloud technology, modern software, and AI can be game changers for exactly the kinds of difficult problems that CEO’s — and an entire population newly attuned to their dependence on the supply chain — have grown increasingly impatient about fixing.
In the next Blog in this series, we’ll dive right into DATA, and talk about the crucial role its quality and availability plays for Supply Chain solutions leveraging AI. Stay tuned!