Written by Marijn Overvest | Reviewed by Sjoerd Goedhart | Fact Checked by Ruud Emonds | Our editorial policy

ChatGPT for Category Management: A Practical Workflow

The Category-Management Job, and Where AI Fits

Category management is the procurement discipline of understanding a spend category deeply enough to drive value beyond price. Market dynamics, supplier landscape, demand patterns, internal stakeholders, regulatory and sustainability context, all of it informs the sourcing strategy, the supplier panel, and the negotiation positions. A category manager who knows their category is irreplaceable; a category manager who only manages contracts is automatable.

AI tools have not changed what category management is. They have changed how much category depth a single category manager can credibly maintain. The 8-10 hours per week previously spent on market research, supplier research, internal data gathering, and deck production can collapse to 2-3 hours, freeing capacity for the work AI cannot do, supplier relationships, stakeholder conversations, judgement calls on category strategy.

ChatGPT in 2026 is the most accessible AI tool for category management, with the broadest user base and the most familiar interface. For category managers who have not yet started using AI seriously, ChatGPT is the easiest entry point. The 2026 AI Readiness in Procurement survey found that 61% of procurement teams use ChatGPT or GPT-4, the highest adoption of any AI tool. The familiarity matters; the capability is real.

Five ChatGPT Use Cases that Change Category-Management Work

These are the five ChatGPT applications that consistently produce value for category managers in 2026. Each replaces a recurring multi-hour task with a 15-30 minute prompt-and-refine workflow.

1. Market context briefs

Before any major sourcing event or category review, the category manager needs a current view of the market: who the major suppliers are, recent M&A activity, pricing trends, regulatory developments, sustainability and supply-chain risks. ChatGPT with web search produces a structured market brief in 15-20 minutes. The output is not a substitute for years of category knowledge; it is a starting reference that frees the senior category manager from re-researching basics.

2. Spend deep-dive narrative

Drop the category's spend file into ChatGPT and ask for a structured narrative: top suppliers, year-over-year change, category share, growth patterns, anomalies worth investigating. The output is a four-paragraph narrative that frames the category review deck. Hours of pivot-table work collapse to one prompt.

3. Supplier shortlist research

For new sourcing events, the category manager needs a credible supplier longlist. ChatGPT with web search produces a structured shortlist with each supplier's profile, capabilities, customer references, geographic footprint, and any red flags. Manual due-diligence still happens, but on a curated list rather than a blank-slate search.

4. Tender document drafting

RFI, RFQ, and RFP documents follow standard structures with category-specific content. ChatGPT drafts the structural skeleton and the boilerplate, including evaluation criteria, response format, and timeline. The category manager refines the category-specific content. Drafting time drops from a full day to a half-day, with the standard structure carried forward.

5. Category strategy narrative

The annual category strategy document, what the category is, where it is going, what the procurement function will do over the next 12-24 months, is the highest-leverage deliverable a category manager produces. ChatGPT does not write the strategy; the category manager does. But ChatGPT structures the draft, finds the gaps in the argument, and produces the polished narrative version of the analysis. The strategy gets sharper and lands better with executive readers.

The Monthly Category-Management Workflow with ChatGPT

A category manager covering 3-5 categories can sustain the following rhythm with ChatGPT support. The point is consistency; the discipline compounds quarter over quarter.

Week 1 of the month, spend pulse. Drop in the previous month's spend by category. Ask ChatGPT for a 5-bullet summary per category: top three movers, anomalies worth investigating, suppliers approaching contractual thresholds. 20 minutes per category, 90 minutes for a portfolio of five.

Week 2, market and supplier signals. ChatGPT pulls market intelligence on the active categories, sector pricing trends, supplier news, regulatory updates, sustainability developments. Output is a one-page-per-category brief. Used as the input for any conversations with suppliers or stakeholders that week.

Week 3, the deeper dive. One category gets a deeper review each month, rotating across the portfolio. Spend deep-dive, supplier panel review, strategy refresh. ChatGPT structures the analysis; the category manager refines and drives conclusions.

Week 4, deliverables. Whichever deliverables are due this month, category council update, supplier business review prep, savings register update, ChatGPT produces the drafts; the category manager edits and signs off.

The compounding pattern: by month three, the prompts are sharper, the data feeds are cleaner, and the deliverables look like the work of a category manager with twice the bandwidth they used to have.

Worked Example: A Marketing Services Category with ChatGPT

A category manager owns marketing services, EUR 8M annual spend across creative agencies, media buying, MarTech licenses, and production. They run a quarterly category council. The Q3 prep, using ChatGPT, takes a morning instead of the previous three days.

Step 1, spend narrative. Drop Q2 spend file. Ask: "Produce a category narrative for marketing services Q2. Cover: top suppliers, year-over-year change, share of category, growth and contraction patterns, anomalies. 6 paragraphs." Output: a credible draft narrative in three minutes.

Step 2, supplier landscape. Ask: "For each of the top six marketing services suppliers in this list, produce a 4-line profile: capability, recent M&A or leadership changes, customer references in [our sector], any red flags from the last 12 months." Output: structured supplier profiles for the supplier panel slide.

Step 3, strategy options. Ask: "For this marketing services category with the spend pattern above, what are the three strategic options for the next 12-24 months, given the supplier landscape and our growth plans?" Output: three credible strategy options with pros and cons, used as the category council discussion framework.

Step 4, council deck. Ask: "Produce an 8-slide category council deck for marketing services covering spend, suppliers, market, the three strategy options, the recommendation, and the decisions needed." Output: a structured deck the category manager edits down.

Total time: 3 hours including the human edit pass. Previous time: 2.5 days. The category council itself is the same length, but the prep is recovered. Categories without this discipline see strategy decisions slipping; categories with it run on time. Category Management in Procurement Course covers the category-strategy discipline this workflow plugs into.

Limits and When to Switch to a Different Tool

ChatGPT has clear limits for category management work. Three patterns where another tool is the right choice.

For long contract analysis, Claude is often the stronger option. ChatGPT handles short documents well but degrades on 60+ page MSAs in a way Claude does not. Category managers handling complex MSAs frequently use both: ChatGPT for the rest of the workflow, Claude for the contract review step.

For document-heavy work where files live in SharePoint or Google Drive, Copilot or Gemini's native integration with the respective ecosystem reduces friction. ChatGPT requires upload; Copilot and Gemini already see the files.

For real-time spend analytics over millions of rows, a proper analytics tool, Power BI, Tableau, or a spend-management platform, beats any general AI for the visualisation and refresh layer. ChatGPT helps with the narrative; the visualisation belongs in the analytics tool.

Governance and How the Team Works with the Tool

Three governance decisions that procurement teams using ChatGPT consistently make.

Single shared workspace, not personal accounts. ChatGPT Team or Enterprise gives the procurement team a shared workspace where prompts, Custom GPTs, and conversation patterns can be shared. Personal accounts produce personal productivity; shared workspaces produce team capability.

Standardised Custom GPTs for recurring work. Each category's recurring deliverables, the monthly spend pulse, the quarterly council deck, the supplier QBR prep, lives in a Custom GPT that the whole team can invoke. The team's standards for these deliverables are encoded once and reused. Custom GPTs for procurement covers the Custom GPT build pattern.

Data-handling policy alignment. The team's AI policy should explicitly cover what category data goes into ChatGPT, what doesn't, and how. Supplier names and spend values typically yes; supplier-confidential commercial terms typically no without contractual consent. Set this up before the first sensitive query.

Common Mistakes that Make ChatGPT Category Work Feel Thin

Asking generic questions

"Tell me about the marketing services category" produces generic output. "Produce a category narrative for [our] marketing services spend [paste data], structured as [structure], highlighting [specific items]" produces output the team can use. Specificity is the difference.

Not feeding the team's data

ChatGPT working from public knowledge produces public-knowledge insights. ChatGPT working from the team's spend file, supplier list, and contracts produces team-specific insights. The data is what makes the output proprietary.

Treating ChatGPT output as the deliverable

The category manager's name is on the deliverable. ChatGPT produces drafts; the category manager edits, refines, and signs off. Treating ChatGPT output as the final deliverable is what produces the 'thin' feeling that experienced category managers spot immediately.

Not building the institutional library

Category managers who keep their best ChatGPT prompts in their head reproduce the work each time. Teams that maintain a shared prompt library (Custom GPTs, Project instructions, documented patterns) build compounding capability. The library is the team's intellectual property.

Frequently asked questions