Leverage ChatGPT for Procurement (2025 Update)
Introduction
In procurement, staying ahead means continuously embracing tools that boost efficiency, insight, and agility. Since our 2024 overview of ChatGPT in sourcing, the underlying technology and prompting techniques have matured rapidly. This update distills recent non-technical explanations of how ChatGPT works, beginner-to-advanced prompt frameworks, and practical tips for buyers to get started immediately.
Contents…
How ChatGPT Actually Works
Even for non-techies, it helps to grasp the “big picture” of large-language models (LLMs) like ChatGPT:
- Training on Token Statistics
- ChatGPT analyzes vast corpora of text, learning which sequences of “tokens” (words or word-pieces) tend to follow one another.
- Transformer Architecture
- A “transformer” network weighs every token’s relationship to every other token in a prompt, enabling remarkably coherent generation (Berner, 2025).
- Fine-Tuning and Safety Layers
- After initial pre-training, the model is refined on domain-specific examples and overlaid with safety filters to reduce hallucinations and bias (Berner, 2025).
Key takeaway: You don’t need to code or understand algorithms to use ChatGPT—just think of it as a highly knowledgeable, context-aware writing assistant.
Prompt Engineering: From Basics to Mastery
1. Beginner Prompts
Jodie Cook (2025) outlines ten prompts that guide you from “list five” up through basic refinement—ideal for first-time users:
- “List five criteria for evaluating a new supplier”
- “Summarize this RFP into key objectives”
- “Rephrase this email to sound more persuasive”
(Cook, 2025)
2. Intermediate Techniques
Sheetal Bandari (2025) recommends ten “easy ways” to level up prompt precision:
- Be specific: define input formats (“Output as a table with columns: Supplier, Score, Risk”).
- Use examples: provide sample inputs and expected outputs.
- Set the context: “You are a procurement analyst specializing in electronics.”
(Bandari, 2025)
3. Advanced Strategies
- Chain-of-Thought: ask the model to “think out loud,” improving complex reasoning.
- Role-Playing: “Act as an experienced negotiator and propose opening offers.”
- Iterative Refinement: feed back ChatGPT’s own answers for further polishing.
(Cook, 2025; Bandari, 2025)
Getting Started as a Buyer
- Choose a Task: e.g., supplier research.
- Draft a Simple Prompt: “List top 5 electronics suppliers in Southeast Asia, with annual spend > $10M.”
- Refine: Add format instructions—“Present results as a CSV table.”
- Iterate: Ask, “What additional data would strengthen my supplier shortlist?”
Tip: Save your best prompts in a shared “Prompt Library” for your team.
Applying ChatGPT Across Procurement
| Use Case | Example Prompt |
|---|---|
| Category Strategy | “Analyze 2024 spend data for indirect materials and identify three cost-reduction opportunities.” |
| Supplier Discovery | “Find tier-1 steel suppliers in Europe with ISO certification; output as a table with contact details.” |
| RFP Drafting | “Generate an RFP outline for IT services, including scope, SLAs, and evaluation criteria.” |
| Negotiation Prep | “Role-play as a supplier; propose behalf arguments for a price increase and suggest counterarguments.” |
Best Practices & Next Steps
- Validate Always: double-check critical facts (prices, certifications).
- Maintain Data Privacy: avoid uploading sensitive PII into any AI tool.
- Continuously Learn: subscribe to AI-in-procurement newsletters and update your prompts quarterly.
By adopting structured prompting strategies—from simple lists to advanced chain-of-thought dialogs—buyers can unlock ChatGPT’s full potential in every procurement phase.
A Learning Journey: Mastering ChatGPT for Procurement
Module 1: Define Your Needs & Features
Objective: Learn to translate business requirements into clear, actionable prompts.
- What to do: List out the key technical or commercial attributes for your target product or service.
- Example (iPhone 12):
• Processor: A14 Bionic • Network: 5G support • Display: Super Retina XDR • Camera: Advanced multi-lens system • Extras: Wireless charging, sustainability criteria
- Your exercise: Replace “iPhone 12” with your own category (e.g., “industrial pumps”) and list 4–6 “must-have” specs plus business drivers (pricing, lead-times, ESG).
Module 2: Ask Smarter, More Sophisticated Questions
Objective: Move from simple lists to rich, targeted comparisons.
- Begin with basic feature queries:“Which smartphones with A14-level processors have sustainable supply chains?”
- Level up with multi-factor prompts:“What 5G smartphones in 2024 match iPhone 12 display quality and camera specs, priced under $800?”
- Compare broader procurement factors:“Which suppliers offer reliable global distribution for these models?”
- Your exercise: Create two multi-criterion prompts that combine product specs + supplier/region/pricing.
Module 3: Market Research & Trend Analysis
Objective: Harness ChatGPT’s ability to surface competitive intelligence and forecasts.
- Trend spotting:“What emerging smartphone features should buyers watch in 2025?”
- Supplier benchmarking:“Which brands lead in eco-friendly handset manufacturing?”
- Pricing outlook:“Forecast average retail prices for mid-range 5G phones through Q4 2025.”
- Your exercise: Ask ChatGPT for three supplier-side trends in your category and summarize in a short bullet list.
Module 4: Build Data-Driven Strategies
Objective: Turn research into clear deliverables—reports, RFQs, negotiation plans.
- Report drafting:“Compare smartphones vs. competitors on price, specs, and sustainability; output as a table.”
- RFQ template:“Draft an RFQ for 5G devices emphasizing cost-efficiency, ESG, and delivery KPIs.”
- Negotiation prep:“Act as a supplier: propose a price increase rationale; then suggest buyer counteroffers.”
- Your exercise: Use ChatGPT to generate an outline for one key deliverable (e.g., an RFQ or negotiation memo).
Module 5: Verify, Validate & Keep Current
Objective: Instill a habit of fact-checking and staying up to date.
- Cross-check data: Always compare ChatGPT outputs with primary sources (manufacturer sites, industry databases).
- Live update prompts:“What reliable sites track real-time pricing and reviews for [your category]?”
- Your exercise: Create a “verification prompt” you’ll use whenever ChatGPT gives you hard figures (prices, lead times, certifications).
Module 6: Unlock Productivity Gains & Advanced Workflows
Objective: Automate routine tasks and integrate ChatGPT into your daily toolkit.
- Supplier discovery: Ask for niche or regional vendors you might miss.
- Automated summaries: Feed procurement spreadsheets into ChatGPT (via API or Apps) to get executive-ready dashboards.
- Contract review: Summarize key clauses and flag risks in standard T&Cs.
- Continuous learning: Schedule a weekly “Ask ChatGPT” check-in on regulatory or market shifts.
Ready for even more?
Enroll in Intro to ChatGPT Apps for Buyers to learn how to connect ChatGPT to your own data sources, build internal apps, and automate procurement workflows end-to-end.
This structured path takes you from defining what you need, through asking ever-more powerful questions, all the way to embedding ChatGPT into your everyday procurement processes. Start with Module 1 today, and see how quickly you’ll unlock new insights and efficiency gains.
When using ChatGPT or similar AI tools, especially for business applications like procurement, there are important risks associated with handling confidential or sensitive data. Here are key concerns and best practices for mitigating these risks:
Dangers of Using Confidential Data with ChatGPT
- Data Exposure Risk:
AI models like ChatGPT process data by sending it to servers for computation, which poses a potential risk of sensitive information being exposed during this process. This can include proprietary product data, supplier contracts, or financial details that could be accessed by unauthorized parties. - Third-Party Data Handling:
OpenAI (and similar organizations) may log or store interactions with the model, depending on the specific terms of service. If you’re inputting sensitive data (e.g., supplier negotiations, pricing models), that information could be stored in logs, increasing the risk of a data breach. - Lack of Encryption:
ChatGPT sessions may not always employ encryption standards comparable to secure systems designed for handling confidential information. Even though the connection might be encrypted, the data is still processed by external servers, increasing the risk of interception or misuse. - Inadvertent Data Sharing:
AI tools can create responses based on previous conversations with other users, which means there is a risk of cross-user data leaks. While AI providers work to prevent this, it is possible that sensitive information might inadvertently be used in training or model improvements if shared without the proper safeguards.
Best Practices to Avoid Risks
- Avoid Inputting Confidential Data:
Do not input sensitive data such as supplier names, pricing, contract terms, personal information, or trade secrets directly into ChatGPT. Stick to generic queries or scenarios when using AI for procurement purposes. - Use Anonymized Data:
If you need to use ChatGPT to simulate a scenario or draft reports, anonymize or abstract any sensitive information. For example, instead of using real product names or prices, replace them with placeholders or general terms. Example: Instead of inputting “negotiating a 5% discount with Supplier X for $1M order,” use “negotiating a percentage discount with Supplier Y for a large bulk order.” - Follow Company Policies:
Many organizations have specific guidelines regarding the use of external AI tools to handle sensitive data. Always adhere to your company’s data security and compliance policies when using tools like ChatGPT. - Use Private or On-Premise Models:
If your organization relies on AI to handle sensitive information, consider using an on-premise version of the model or a secure, enterprise-level version of ChatGPT. These versions offer more control over data privacy and storage, ensuring confidential information remains within your company’s secure infrastructure. - Consult Legal and Compliance Teams:
In regulated industries (e.g., finance, healthcare, or defense), inputting confidential data into third-party AI systems may violate privacy regulations such as GDPR, CCPA, or industry-specific laws. Always consult your legal or compliance team before using AI for tasks involving sensitive information. - Understand Terms of Use:
Review the terms of use of the AI tool you’re working with, ensuring you are fully aware of how your data will be processed, stored, and handled. Some AI providers may offer guarantees or assurances about data privacy, while others may not.
By maintaining awareness of these risks and adopting safe practices, you can leverage the benefits of AI tools like ChatGPT without compromising confidentiality or security in procurement processes.
Conclusion: Leverage ChatGPT and increase your productivity.
ChatGPT has evolved into a powerful tool that procurement professionals can rely on for everything from product research to supplier discovery and contract management. By embracing its capabilities, you can streamline operations, make more informed decisions, and stay agile in a rapidly evolving marketplace.
If you find competing products, you’ll also find competing manufacturers and suppliers that can be recipients of your RFQ. Learn more about the sourcing process in our Sourcing Engine courses.
Are you interested in sharing successful scripts (instructions) to fellow buyers?
Are you navigating the exciting world of AI and exploring how to leverage ChatGPT? LHTS believe that your expertise and experiences are invaluable, and we would love for you to share them with a community of like-minded professionals.
LHTS are reaching out to invite you to contribute to an innovative initiative: sharing successful scripts and instructions on how professional buyers can use Chat GPT. Your insights could help others streamline their operations, find new efficiencies, and make more informed decisions.
By participating, you will not only be aiding your peers but also gaining access to a curated collection of strategies and approaches shared by other professionals in the field. This is a fantastic opportunity to learn from each other and stay ahead in the ever-evolving landscape of procurement and AI technology.
If you are interested in sharing your successful scripts, LHTS will convert your scripts to explanatory blogposts and create value to the procurement community. More information how to contact us here.
References
Bandari, S. (2025). 10 easy ways to master prompt writing for ChatGPT. Retrieved from LinkedIn Pulse.
Berner, K. (2025). A non-techy take: How chatbots like ChatGPT work. Retrieved from LinkedIn Pulse.
Cook, J. (2025, April 29). 10 ChatGPT prompts to go from beginner to advanced at prompting. Forbes.
Note: Illustration to the blogpost “How Professional Buyers Can Leverage ChatGPT for Procurement Efficiency in 2025” was created on September 14, 2024.