I interviewed Richard Barnett who discussed The Role of AI in Strategic Sourcing.







Great to speak with you today, Richard, and looking forward to this topic on the role of AI in strategic sourcing. Can you first provide a brief background of yourself?


You bet, Dustin. I am senior vice president of marketing and customer success for LevaData. I have worked over 20 years in the supply chain solution provider space starting off my career with i2 Technologies in the late '90s, played different roles with supply chain innovators, focused on supply chain planning and optimization, deployment of business networks, particularly in supply chain procurement and logistics, and then more recently with my role at LevaData, where we're really focused on strategic sourcing and procurement.


Can we first start by defining what is AI?


I think this is an interesting topic. You'll hear AI being used in different contexts related to automation or the threat of AI in terms of this notion of the robots will take over human jobsbecause of the idea of intelligent computers that can replicate, or near, the cognitive abilities of the human brain. The real story, though, that we don't really understand oftentimes that's evolved over time is AI is a discipline is an area that really goes back to the 1960s and efforts in early computing and robotics to replicatepieces of what the human brain does. It could be in terms of sensing, the five senses. It could be in terms of repeating what humanscould do from a manual process perspective in robotics, and it could be in terms of calculating or predictingthe future based on information processing to understand patterns, for example.


And today, what we have is a combination of technologies which are broadly put under the AI umbrella, but is best defined as “Narrow AI”, as they are applied in specific contexts.


At LevaData, when we think about AI, we really look at more specifically the areas of machine learning, pattern recognition, particularly with respect to new data sources that are now available. So how do you stream and derive intelligence from massive large datasets?But also optimization and predictive technologies implemented at scale with cloud computing or new algorithms that we can apply to specific situations where you're trying to predictcost trends, for example.



Can you help the audience understand how AI will impact strategic sourcing professionals?


It's a great question. What we're seeing is the early period of adoption in a very pragmatic way of AI-based capabilities broadly in supply chain and then within strategic sourcing. Unfortunately, strategic sourcing and direct materials procurement is even further behind the innovation adoption curve. So our focus is on enabling Cognitive Sourcing, our term for describing this intersection of applied AI technologies for the purpose of augmenting the intelligence of supply chain and strategic sourcing professionals to make better decisions. The foundation is based on a three-level framework of first, gathering all of the best information and data sources from within the enterprise combined with market intelligence, with second, identifying the best actionable insights from this constantly updated information to third, optimize decisions and negotiation outcomes. Most companies today are relying on very fragmented data sources that relate to the enterprise history of spend and bill of materials (BOMs) and product information.And trying to marry that up with the very robust but very diverse set of information that's more market intelligence or external to the enterprise and bringing those two worlds together is really the foundation.


And then on top of that, we apply different AI technologies, to derive insights around where emerging risks or opportunities are, and it's always on. It is constantly looking and trying to correlate what the impact of, say, a change in demand for a certain product is internally, maybe a change in the forecast for the holiday season. What does that mean for the impact all current areas of spend, the suppliers that are going to be dependent on delivering that increase in volume, and maybe where there additional cost savings? What potential emerging risks because of the capacity change in the market?


So we filter those insights into what we call risks and opportunities. And then we have an intelligent agent called Leva, which is a combination of both intelligent assistance as well as AI cognitive sourcing technology that deliver recommendations to professionals are what the most important risk and opportunities are to act upon, what negotiation levers might be useful to use, in what sequence, and at what time to negotiate for either better price or reduced risk with each supplier.


And then it learns over time based on those iterations of moves that companies make to constantly improve and get better.This includes the accuracy of predicting costs or risk changes or the combination of negotiation levers that are used to produce a cost savings— not just across an individual customer, but actually across the community of all LevaData users and customers. So it's an interesting combination of looking at predictive optimization models, chatbot-based intelligent assistance, in addition to machine learning and crowd-sourced intelligence.


So those are the set of AI technologies that we believe are the fundamentals of what we call this new area of Cognitive Sourcing.


Can you explain the cognitive sourcing a little bit more? And also, you mentioned earlier that you have a recent study that you did.


You bet. I think at a high level, it's really a shift from interacting with the market maybe with annual or quarterly major sourcing events where companies have traditionally not really changed their process or their strategy in these areas for 15 to 20 years, really. One of the challenges a lot of these companies have today is that they can't act upon the entire long-tail of the spending that they're managing. They've got a complex set of suppliers or categories or sub-commodities of spend. They have a hard time actually covering, say, 10,000 suppliers in all the spend categories. They can't scale out across the opportunity.


The second problem is that they're not oftentimes very proactive around predicting changes that might occur in the market and acting ahead of their competitors. So they're falling in the same cycle where there is maybe a big supply constraint that's occurring in memory or recently in electronics resistors, for example, which is not individual piece-part price for resistors. It's very, very low. It's sometimes less than 0.01 cent. But if they buy a billion of them, it's really meaningful at volume and scale. The approach with cognitive sourcing is to — I call it — flipping the classroom. It's giving all the information that you need in a filtered, in-context way to a sourcing professional so that they can actually act in the market in what we call more of a continuous sourcing cycle.


So instead of actually only going to the market once a year or once a quarter with information that may be two or three months old, what if you had the capability that was supporting you with recommendations, giving you insights. Then the net result is that you're scaling and looking more comprehensively across most aspects of your inbound supply chain, including costs and risks, and then, more importantly, you're able to take action proactively to avoid risks and challenges in the market, particularly in markets that are very competitive. That becomes a competitive advantage for these companies.


So we did a study that we are just about the complete the final published version of it, but we launched last week at the first Cognitive SourcingSummit in Palo Alto. The results of the study were pretty interesting because on one hand, 93% of the survey respondents said that they want a data-driven capability in their procurement organization, and they're very interested in AI to sense these risks and opportunities or to improve negotiations. But only 52% believe that the talent on their teams are actually ready to leverage and use this information and ready for a digital transformation, in short.


We also found that 48% really want a market-intelligent system that combines information from multiple sources and presents recommendations. But a surprisingly 38% said that they don't need market intelligence; they get what they need from their supplier directly, which is very interesting because it sort of indicates this kind of schism in the market where companies are, maybe on the one hand, comfortable with the relationships that they have with their suppliers where they're asking their suppliers to give them the information that they need.


And when we discussed this survey during the conference last week, Angel Mendez, who was one of the guest speakers, said in the sourcing profession, it's like they're “bringing a knife to a gun fight” because of the investment, sophistication of the supplier side or the seller side of the dynamic has increasingly gotten more sophisticated. They're seeing all the market information, and they're going into these negotiations much better armed than the actual buyers are, particularly in strategic sourcing and procurement of direct material sourcing teams in large global manufacturers.


And so the idea is to level the playing field and not rely on the supplier information but actually provide new sources information and insight.


The other aspect of the survey that was pretty interesting is that we looked at the RFQ process, particularly around direct material sourcing. What we found was that only 56% of all suppliers are actively engaged annually out of 100% of the total supply base. And they're using, oftentimes, 55% email and spreadsheets or custom development tools to manage that RFQ process.


We also found that the average performance from the RFQ cycle time is 14 days in preparation and 26 days in negotiation, taking a little bit over 40 days to complete a cycle. But we found some best-in-class companies that are operating on one day of preparation, two days of negotiation, and three days total cycle time.


So our observation is that that is a massive opportunity for competitive advantage for companies to actually change and look at their model very strategically around how they're engaging with their supply base and use these cognitive technologies to get them to a best-in-class performance. And you can just kind of think about it in terms of how two major competitors might be working in the market. The one that's actually able to work in days versus weeks and months is going to win, particularly in these markets in high-tech and consumer electronics that are fast-moving product lifecycles, very tight margins and price curves that change in a matter of months. That becomes pretty mission-critical to look at these areas and a new source of innovation.


But the big takeaway we had at the conference was everyone agreed that this area has just not received the right investment in innovation and support that maybe other aspects of supply chain or other groups and departments within these companies have experienced.  My point of view is that I think we should really look at this area and really prioritize it look for innovation and really advocate for a conversation to explore the art of the possible because the technologies and the capabilities are just landing now. And they're potentially highly disruptive for early adopters who really kind of take this as part of their key strategy. So that's what's pretty exciting about this space.


Thanks, Richard, for sharing today, and I look forward to staying in touch if you want to continue discussions on this topic.



About Richard Barnett






Richard Barnett


SVP Marketing and Customer Success


LinkedIn Profile