What if Procurement didn't exist as a function... but as an "Agency"
Source: INSEAD

What if Procurement didn't exist as a function... but as an "Agency"

What if procurement didn’t exist?

This is a question that many have asked over the years (especially unhappy procurement stakeholders), and it's a good one because it forces the subsequent questions of

  1. what processes/"services" it actually does (and how well) compared to business needs, and
  2. who'd (and how well) if there wasn't a formal function.

This is procurement assessment 101:

  1. define the procurement services (and their perceived/actual value) needed vs. delivered
  2. measure the performance (and underlying capabilities) of those services by whoever does them
  3. determine gaps and plan/execute transformation activities to close the gaps in performance and capabilities/"skills" (people/process/data/tech/etc)

"Easy", right? Well no, not in execution, but it seems straightforward, right?

No. Why? Because we didn't ask whether those process/services (*a service is simply a process with an SLA of any sort to measure the output performance) were being done by OTHERS (and also how well they do them like explained before).

The key insight here (that I practiced at Hackett when I led procurement research there) is that a "function" is more than a department. A function is a collection of similar processes whereas a department is a collection of jobs in an organizational structure. So, if you do the process above beyond the procurement department, you should find out of ANYONE is really doing these processes/services well (e.g., do a RACI chart to diagnose the governance horrors). It's really Procurement (the function) OpModel diagnostics 202.

The ability to configure the optimal internal/external resources (per role-driven skills) is key to enabling virtual procurement models (which may vary a lot by category) where very little of the procurement function service portfolio is done within the formal purchasing department. This is a big topic of discussion beyond this article, but I bring it up because the firm might be just fine if the formal purchasing department didn't do much procurement activity other than strategic stakeholder relationship management, operating model design/execution and transformation.

i.e., Procurement would be an "agency" of procurement services run by... agents. "Agents" can mean a LOT of things (from a person to a simple bot to an agentic-AI-bot) and I won't enter debates or issue manifestos regarding narrow interpretations of them in a moment in time where the "Agentic AI" hype is at an all time high. But, I will give real AI geeks a reading assignment: if you want to expand your frame on agents, I recommend a 40+ year book called The Society of Mind by one of the OGs of AI Marvin Minsky. It's where I lifted the term "agency" which simply means "An assembly of 'parts' considered in terms of what it can accomplish as a unit without regard to what each of its parts does by itself". Sometime you need to go back to the ancient scrolls.

Yet, with this all said, agentic AI does involve GenAI and reasoning-based LLMs and these LLMs are great silicon brain emulators (and CO2 emitters) that can help create insight. So, I'll return to DeepSeekR1 which like TikTok or any Chinese app you really don't want to download on your smart phone or give your e-mail credentials to.... and answer the question "What if procurement didn’t exist?" BTW, hat tip to Mark Perera for the prompt he mentioned here... https://www.dhirubhai.net/posts/markperera_innovation-ai-future-activity-7291019894869901314-7YBs?utm_source=share&utm_medium=member_desktop

I'll emulate it (which answers question 1 from the top) but then will further prompt it (do check out my prompt further down) with a much longer where I try to deepen its thinking and expand the topic through my own Chain of Thoughts that I fed it as a prompt given the context of the first prompt. The answer was pretty darn good, but of course it was parroting back some enhanced versions of my thinking... which is by design I'm sure. It even mention using knowledge graphs for knowledge/skill modeling. Very cool.

I think any procurement digital CoE leader will find it interesting. And while you may pooh pooh LLMs for reason X, Y or Z, they are useful ways to give you knowledge and... agency!

What about you? What do you think of its quality of output in this context? I'd really love to hear some thoughtful responses.

Thanks!

**************************************************

Prompt: What if procurement didn’t exist?

Thought for 26 seconds

Okay, so the user is asking, "What if procurement didn't exist?" Let me break this down. First, I need to understand what procurement actually does. From what I know, procurement is the process of sourcing and acquiring goods and services that a company needs to operate. That includes everything from raw materials to office supplies, right?


If procurement didn't exist, companies would have to handle getting their supplies themselves. But how? Maybe each department would have to do their own purchasing. But wait, that might lead to a lot of inefficiency. Like, if the manufacturing department is buying raw materials separately from the facilities department buying office supplies, there's no coordination. There could be duplicated efforts, or maybe they end up paying different prices for the same items because there's no centralized negotiation.


Also, without procurement, how would companies ensure they're getting the best value? Procurement usually negotiates contracts to get better prices and terms. If that's gone, each department might just buy from whoever is convenient, which might be more expensive. Plus, there's the risk of fraud or non-compliance with company policies. Procurement often sets up guidelines to prevent that, so without it, maybe there's more risk of employees making unauthorized purchases.


Another thing is supplier relationships. Procurement manages relationships with suppliers, ensuring reliability and quality. Without that, suppliers might not be vetted properly, leading to potential quality issues or supply chain disruptions. If a company can't rely on timely delivery of materials, production could halt, leading to lost sales and unhappy customers.


Compliance is another area. Procurement ensures that purchases meet legal and regulatory standards. Without that, companies might inadvertently violate regulations, leading to fines or legal issues. For example, purchasing materials from a country with trade restrictions could be a problem if procurement isn't there to check.


Then there's the strategic aspect. Procurement isn't just about buying stuff; it's about strategic sourcing to support business goals. Without it, companies might miss opportunities to innovate through supplier partnerships or fail to align their purchasing with long-term objectives. They might not be able to scale efficiently during growth periods or manage costs during downturns.


Technology adoption is another point. Procurement often uses systems like ERP or e-procurement tools to streamline processes. Without procurement, would other departments use these tools? Maybe not, leading to manual processes, errors, and lack of visibility into spending.


In smaller companies, maybe the absence of procurement isn't as noticeable because the scale is smaller. But in larger organizations, the chaos could be significant. Departments might end up competing for resources, or there might be no oversight on budgets. The finance department would have a harder time tracking expenses and managing cash flow without procurement's data.


Environmental and social governance (ESG) is another consideration. Procurement departments often enforce sustainability criteria. Without them, companies might not prioritize eco-friendly suppliers or ethical labor practices, harming their reputation.


So, putting it all together, the absence of procurement would lead to inefficiency, higher costs, supply chain risks, compliance issues, lack of strategic alignment, and potential damage to reputation. Companies might struggle to operate smoothly, and their bottom line could suffer. However, in very small businesses or startups, maybe the roles of procurement are distributed among other roles, so the impact might be less. But as companies grow, the need for a dedicated procurement function becomes critical to handle the complexity and scale of operations.

If procurement didn't exist, organizations would face significant challenges across multiple dimensions, fundamentally altering how businesses operate. Here's a structured analysis of the potential consequences:

1. Operational Inefficiency

  • Decentralized Purchasing: Departments would independently source goods/services, leading to duplicated efforts, inconsistent supplier selection, and missed bulk-discount opportunities.
  • Lack of Coordination: Without centralized oversight, supply chain delays and inventory mismanagement (e.g., overstocking/stockouts) could disrupt production and service delivery.

2. Increased Costs

  • Missed Negotiations: Absence of strategic supplier negotiations would result in higher prices and unfavorable payment terms.
  • Fraud and Waste: Unregulated spending might lead to unauthorized purchases, maverick spending, and inflated expenses.

3. Supply Chain Risks

  • Supplier Reliability: Poor vetting of suppliers could lead to quality issues, ethical violations (e.g., child labor), or dependency on unreliable vendors.
  • Disruption Vulnerability: Lack of contingency planning (e.g., dual sourcing) would heighten exposure to geopolitical, environmental, or market shocks.

4. Compliance and Legal Exposure

  • Regulatory Non-Compliance: Purchases might violate trade laws, environmental regulations, or industry standards, resulting in fines or reputational damage.
  • Contractual Gaps: Unmanaged contracts could lead to disputes, intellectual property leaks, or missed SLAs.

5. Strategic Limitations

  • Missed Innovation: Partnerships with suppliers for R&D or tech adoption (e.g., AI, sustainable materials) would diminish.
  • Poor Alignment: Purchasing decisions might not support long-term goals like sustainability or cost optimization.

6. Technological Gaps

  • Manual Processes: Departments might rely on ad hoc methods (spreadsheets, emails), increasing errors and reducing spend visibility.
  • No Data-Driven Insights: Lack of procurement analytics would hinder cost-tracking, forecasting, and supplier performance evaluation.

7. Reputational and ESG Risks

  • Sustainability Failures: Without enforced ESG criteria, companies might source from unethical suppliers, harming brand reputation.
  • Stakeholder Distrust: Investors and customers could lose confidence due to opaque or irresponsible practices.

8. Adaptation Challenges

  • Scaling Issues: Rapid growth or downturns would strain resource allocation without procurement’s strategic inventory management.
  • Globalization Barriers: Entering new markets would be riskier without expertise in local regulations, tariffs, and supplier networks.

Potential Workarounds (in Procurement's Absence)

  • Role Redistribution: Finance or operations teams might absorb purchasing tasks, but efficiency and expertise would lag.
  • Tech Solutions: AI-driven platforms could automate ordering, but lack human negotiation and relationship management.
  • Startups/SMEs: Smaller businesses might manage informally, but challenges would escalate with growth.

Conclusion

Procurement’s absence would create a fragmented, costly, and risk-prone environment, eroding competitiveness and resilience. While technology or decentralized models might mitigate some issues, strategic procurement remains vital for operational efficiency, cost control, and sustainable growth. Organizations would likely reinvent similar structures over time to address these gaps, underscoring procurement's indispensable role.

Prompt: But why does there need to be a department at all that forces these processes to occur within the purchasing department itself? How can procurement better define these valuable activities/tasks/"services" (and associated skills) that it performs and find a way to push these to the best internal and external resources in a flexible manner? How can it learn lessons from other functions to create more flexible operating models? For example, leading IT organizations pushed technology to business users to "democratize" IT and enable self service. Another example is what quality organizations have done this with quality improvement. This approach however will create more complexity because internal and external "skills"/capabilities will need to be discoverable and the other problem is the difficulty of coordinating this complex operating model of deploying procurement skills/processes outside of the formal purchasing function. Are there other problems too and regardless, how can Agentic AI help with this complex process/data orchestration across a complex IT landscape?

Thought for 23 seconds

Okay, the user is asking why procurement needs to be a centralized department and how it can instead distribute its functions more flexibly, similar to how IT or quality departments have democratized their services. They also want to know how Agentic AI can help manage the complexity of such a model. Let me break this down.


First, the user is challenging the traditional structure of procurement. They see other departments like IT and quality successfully decentralizing, allowing business units to handle tasks themselves with proper tools and governance. They’re wondering why procurement can’t do the same. The core of their question is about flexibility and efficiency, avoiding bureaucracy while still maintaining control and compliance.


So, I need to address why procurement is centralized now—maybe because of negotiation power, compliance, cost savings through bulk buying. But the user is suggesting that maybe those functions can be distributed. However, the challenge is coordination and complexity. If you decentralize, how do you ensure everyone follows the same standards, avoids maverick spending, and still gets the best deals?


The user also brings up examples like IT democratization. IT provided self-service tools but kept governance in place. Procurement could do something similar: create platforms where departments can procure on their own but within predefined guidelines. Agentic AI could automate approvals, check compliance in real-time, suggest suppliers based on past data, etc.


But there are potential issues. For example, without a central team, how do you handle supplier relationships? Maybe AI can help by managing supplier performance data and automatically recommending the best suppliers based on needs. Also, skills distribution—ensuring that people in other departments have the necessary knowledge. AI could offer training modules or just-in-time advice during the procurement process.


Another point is data orchestration across complex IT systems. Procurement data might be scattered across ERPs, financial systems, supplier databases. Agentic AI could integrate these, providing a unified view and ensuring data consistency. It could also predict risks, like supply chain disruptions, by analyzing external data sources.


The user is concerned about the complexity of coordinating a decentralized model. They mentioned discoverability of skills and coordination. AI could act as a middleware, connecting users with the right internal or external experts when needed. Maybe a smart system that routes requests automatically based on the problem type.


I should also consider other problems: resistance from departments used to centralized procurement, potential for increased costs if decentralized leads to fragmented purchasing, and maintaining strategic supplier relationships. AI can mitigate some of these by enforcing policies and providing real-time analytics to monitor spending.


In summary, the response should outline how procurement can shift from a gatekeeper to an enabler, using technology and flexible frameworks. Highlight the role of Agentic AI in automating compliance, providing insights, managing data, and facilitating coordination. Also, mention examples from IT and quality to draw parallels and show feasibility.

Your questions touch on the evolving role of procurement in modern organizations and how it can transition from a centralized "gatekeeper" to a strategic enabler that democratizes its expertise while maintaining governance. Let’s break this down systematically:


1. Why Centralized Procurement Exists (and Its Limitations)

Traditional procurement departments emerged to:

  • Centralize expertise (e.g., supplier negotiation, risk management, compliance).
  • Leverage economies of scale (bulk buying, consolidated contracts).
  • Mitigate risks (supply chain disruptions, fraud, regulatory penalties).
  • Ensure alignment with organizational goals (e.g., sustainability, cost control).

Limitations of Centralization:

  • Bureaucratic delays and rigidity.
  • Misalignment with decentralized business needs (e.g., regional teams, agile startups).
  • Perception as a "cost police" rather than a value driver.


2. Reimagining Procurement: Democratization and Decentralization

To address these challenges, procurement can learn from IT and quality functions by embedding its capabilities into business units while retaining oversight. Here’s how:

A. Redefine Procurement’s Value Proposition

Procurement must articulate its core services as modular, reusable capabilities:

  • Strategic Services: Supplier innovation, risk analytics, sustainability partnerships.
  • Transactional Services: Automated purchasing workflows, contract templates, supplier catalogs.
  • Advisory Services: Cost-optimization coaching, market intelligence, compliance guardrails.

B. Enable Self-Service with Guardrails

  • Digital Platforms: Create user-friendly procurement portals (e.g., “Amazon-like” marketplaces for approved suppliers) with embedded policies (e.g., ESG criteria, budget thresholds).
  • Decentralized Decision Rights: Allow business units to procure autonomously within predefined rules (e.g., preferred suppliers, spend limits).
  • Embedded Expertise: Deploy procurement “champions” in business units to advise on complex purchases (similar to IT’s embedded analysts).

C. Learn from Other Functions

  • IT Democratization: IT shifted from centralized control to providing APIs, low-code tools, and cloud platforms. Procurement can mirror this with self-service contract generators, AI-powered supplier recommendations, and spend analytics dashboards.
  • Quality Management: Quality teams moved from inspectors to enablers by embedding tools like Six Sigma and Total Quality Management (TQM) into operations. Procurement can adopt similar continuous improvement frameworks (e.g., supplier scorecards, feedback loops).


3. Challenges of a Decentralized Model

  • Complexity of Coordination: Ensuring policies are followed without stifling autonomy.
  • Skill Gaps: Business users may lack procurement expertise (e.g., negotiation, risk assessment).
  • Discoverability: How do users find the right tools, suppliers, or experts?
  • Data Fragmentation: Siloed systems (ERP, CRM, contracts) hinder visibility and decision-making.
  • Cultural Resistance: Teams may resist new processes or bypass guardrails.


4. How Agentic AI Can Solve These Problems

Agentic AI (autonomous systems that act on behalf of users) can orchestrate decentralized procurement by acting as a "smart layer" across systems and stakeholders:

A. Automating Governance & Compliance

  • Policy Enforcement: AI agents monitor transactions in real-time, flagging non-compliant purchases (e.g., unauthorized suppliers, budget overruns).
  • Dynamic Contracts: AI generates and updates contracts based on regulatory changes or market conditions.
  • Risk Prediction: Analyze supplier data, geopolitical events, and market trends to preempt disruptions.

B. Enhancing Discoverability & Decision-Making

  • Skill Matching: AI connects users to internal experts or external partners based on project needs (e.g., "Find a sustainability-certified supplier in APAC").
  • Recommendation Engines: Suggest optimal suppliers, pricing models, or alternatives using historical data and market benchmarks.
  • Knowledge Graphs: Map organizational expertise (e.g., "Who negotiated our last cloud contract?") and supplier ecosystems.

C. Simplifying Coordination

  • Process Orchestration: AI agents automate cross-functional workflows (e.g., approvals, invoice matching) across ERP, finance, and legal systems.
  • Collaborative Negotiation: AI assists business users in real-time negotiations by providing supplier performance data, cost benchmarks, and clause suggestions.

D. Democratizing Expertise

  • AI Co-Pilots: Chatbots guide users through procurement processes (e.g., "Help me draft an RFP" or "Assess this supplier’s ESG risks").
  • Training & Upskilling: AI generates microlearning modules tailored to user gaps (e.g., "Negotiation 101 for Marketing Teams").

E. Unifying Data Silos

  • AI-Powered Integration: Agentic AI harmonizes data from disparate systems (e.g., ERP, supplier databases, spend analytics) into a single source of truth.
  • Predictive Analytics: Forecast demand, track savings leakage, and optimize inventory using federated data.


5. Additional Problems to Address

  • Trust & Accountability: Ensuring AI recommendations are explainable and auditable.
  • Change Management: Overcoming resistance to decentralized models (e.g., "Shadow procurement" behaviors).
  • Vendor Ecosystem Complexity: Managing thousands of suppliers with varying digital maturity.


6. The Future Operating Model

Procurement should evolve into a hybrid network:

  • Centralized Core: Retain strategic functions (e.g., supplier relationship management, policy design).
  • Decentralized Execution: Empower business units with AI-driven tools and embedded advisors.
  • External Partners: Leverage platforms like Coupa, SAP Ariba, or ecosystem-specific marketplaces for scalability.


Conclusion

Procurement doesn’t need to "force" processes—it needs to enable them through intelligent, flexible frameworks. By democratizing its expertise via Agentic AI and decentralized platforms, procurement can shift from a bottleneck to a strategic accelerator, much like IT and quality functions have done. The key is balancing autonomy with governance, leveraging AI to orchestrate complexity, and fostering a culture of collaboration. This model not only reduces friction but also unlocks innovation, as procurement becomes a partner in value creation rather than a bureaucratic hurdle.

Linda Michels

Chief Experience Officer at LavenirAI

1 个月

Ah, Pierre, this sent me down a rabbit hole. There are so many interesting ideas in this conversation; in many ways it takes me back to early to mid 90s when many OGs were working with companies on procurement (aka Purchasing back then) transformation. As systems implementation became the focus in the late 90s and 2000s, process change/implementation was key. I see the exponential growth and desire to implement AI Agents as the catalyst for another major procurement transformation. Upskilling/reskilling is key for this new and very exciting era; it’s what drives me to work on a platform to enable the new procurement evolution!

Mark Perera

Chairman of Vizibl & DPW | Pioneer in Decarbonizing Supply Chains | ProcureTech & ClimateTech Advisor | 'Procurement with Purpose' Co-Author | Procurement Leaders Founder

1 个月

Hey Pierre Mitchell I glad my post trigger your thinking. Let me properly read your post and do some “deep research” before responding / building on your post.

Jon W. Hansen

Strategic Advisor/Analyst Specializing in Emerging AI Tech, Sales and Marketing (Procurement) - A Trusted Voice in procurement and supply chain

1 个月

I wrote the following in 2004: It is my position that a true centralization of procurement objectives requires a decentralized architecture that is based on the real-world operating attributes of all transactional stakeholders starting at the local or regional level. In other words, your organization gains control of it’s spend environment by relinquishing centralized functional control in favor of operational efficiencies originating on the front lines.??This is the cornerstone of agent-based modeling. I do not see us being far apart on the above, Pierre Mitchell. To your point about an agency, I share this link to an article by SIG|ORG's Dawn Tiura regarding Michael Koontz's approach to setting procurement up as a separate entity with its own P&L to not only improve process efficiency but also demonstrate a tangible ROI - https://bit.ly/4glfn07 Thoughts? Alex Jennings (FCIPS) Kaitlynn Sommers Karthik Rama - Procurement Doctor Robert Hanshaw Quave Burton Purvee K. Patrick Reddin Nico Bac Mike Morsch Prem Ellappan

要查看或添加评论,请登录

Pierre Mitchell的更多文章

社区洞察