Generative AI: Procurement's New Advisor
Co-Authors: Pierre-André Pancaldi , David (Bryan) Doepken , Pranav S. Lloyd DUFOUR Robert Fuhrmann
The iPhone was released in 2007 and changed the world. The release of the iPhone led to the “iPhone movement” era. Now, Generative AI is becoming just as popular and important. We are currently experiencing the “Generative AI movement” era.
What does that mean for procurement? We know that Generative AI has the potential to change many aspects of a business, particularly procurement. Integrating Large Language Models (LLMs) into the digital procurement landscape can generate insights and content. Generative AI, combined with other technologies, can facilitate faster and easier decision-making and task completion.
Generative AI does this in (2) ways:
This is particularly for areas that do not require creativity or emotional intelligence. Strategic "value-added" capabilities can ultimately drive more accurate procurement decisions.
What is the impact potential of generative AI across procurement capabilities?
There are a growing number of use cases that can support Procurement. Let us examine 3 of them where Generative AI can create a significant impact in the near future.
Use Case 1: Buying Experience
The procurement organization has made significant strides in improving the procurement experience for business users. Generative AI can transform a “purchase request” into a conversation that guides users to the right channels. Generative AI shows us how it takes this experience to the next level.
We all are familiar with TSA pre-check services. Similarly, the Generative AI pre-vets these channels for compliance and policy. The technology can facilitate decision-making and collecting business-wide insights by acting as a category manager.
For example, a Generative AI chatbot places an order for 500 semiconductors with the user to the preferred supplier. The chatbot delivers all factors such as lead times, delivery, risk, quality, and cost in real-time.
Generative AI can provide suggestions to optimize future buys at a better overall value. It contextualizes the request and anticipates the buyer's needs. Generative AI can save users significant time and boost contract compliance by eliminating traditional point-and-click interactions with ERP systems.
It models the right steps of procurement as a “default” behavior.
Use Case 2: Supplier Management and Risk Management
Managing suppliers can be a complex and time-consuming process. But with Generative AI, the process is simplified and accelerated across the entire supplier management lifecycle.
A Generative AI chatbot can centralize communications related to supplier onboarding, provisioning access, and answering questions about the engagement. This technology can reduce onboarding roadblocks and help suppliers understand the company's business needs. The value is captured in faster delivery.
In supplier performance management
Use Case 3: Category Management and Strategic Sourcing
Generative AI improves the processes of defining category plans and sourcing strategies, even with today’s plethora of system applications. Here, we will review some examples.
By absorbing the work of category managers on market intelligence, Generative AI enhances the Category insights dashboard. Generative AI analyzes data from multiple data sets (inside and outside the company) and identifies the best opportunities. The technology provides real- or near-real-time market intelligence and innovation trends for a category's key scopes. It identifies opportunities for the category manager to optimize value, such as identifying price differences across different countries. Generative AI consolidates contracts that cover the same services.
Procurement can strengthen stakeholder and supplier relationships. With the help of Generative AI, procurement can become the cross-functional leader for the business. It tailors analysis for key stakeholders, supports intelligence gathering, identifies suppliers based on capabilities, and augments RFx analysis. It even scores proposals based on the best fit.
Generative AI proactively monitors for risks and proposes mitigation plans. Businesses are affected by these challenges when tension rises in a region where a preferred supplier is located. Instead of being reactive, Generative AI recommends securing upcoming supplies from a different location. Generative AI drafts communication to the new vendor. It suggests any contract changes required and provides a target price with the negotiation strategy. A game-changer for companies that need to make quick sourcing decisions in the face of unpredictable events.
Parting Thoughts
Generative AI has the potential to revolutionize our procurement operations and drive value for businesses. However, realizing its benefits requires careful consideration of critical issues. Here are some key considerations:
• Review your data foundations: Collaborate with leadership to assess your data landscape and its impact on your enterprise's digital roadmap.
• Connect with ecosystem partners: Understand your future state digital landscape, including partner roadmaps that incorporate LLMs for Generative AI capabilities.
• Identify key use cases: Begin identifying use cases and use a test-and-learn strategy before scaling.
• Establish security and governance policies: Prioritize security, compliance, and responsible AI. Align on ethical principles, such as fairness and transparency, and establish a robust governance structure.
The “Generative AI movement” era is already here, and procurement leaders must start thinking about its use now.
Contact us if you want to learn more!
AI for Supply Chain | Co-founder @ Axya
2yHaving a LLM in-between suppliers and procurement specialist is a bit too risky in my opinion. LLMs are creative by design, even with a low temperature parameter. The LLM operator need to double check critically the output to see if it match what was expected. For internal use within a procurement team it’s fine since you can always double check if something look a bit too “creative”. But, suppliers won’t be able to do that. So if they start to get hallucinated facts about the enterprise they are working with or about general fact they should retrieve it can be damaging to the relationship. One venue to fix that up would be to use a mix of knowledge-graph for business critical “fact” about the organization and LLM for the user interaction. This way you get the robustness and creativity at the right spot of the user journey.
Really great article!!!!
Consultant @ Accenture Strategy
2yGreat piece Raj!