By Asmaa Gad | 12 min read
The AI tool works. The team doesn’t use it. Sound familiar? According to Gartner, 70% of digital transformation initiatives fail. Not because the technology was wrong, but because the people side was ignored. In procurement, this pattern is especially common because the teams are busy, sceptical, and deeply attached to their existing workflows.
I have seen teams that bought ChatGPT licences for everyone. Six months later, two people are using it. Everyone else went back to their old process because nobody taught them why or how to change. The tool is the easy part. The change management is where the real work happens.
Here is the playbook.
The 5 Reasons Procurement Teams Resist AI (And What to Do About Each)
“This will replace my job”
The fear: AI automation means redundancy.
The fix: Show that AI handles the 40% of their work nobody likes (data entry, report formatting, basic analysis) so they can focus on the 60% they are good at (negotiation, strategy, relationships). Frame it as an upgrade, not a replacement. Show a concrete example from their own workflow.
“I tried it and it gave me garbage”
The fear: AI is not accurate enough to trust.
The fix: The issue is almost always prompt quality, not tool quality. Run a 30-minute session where you show the same task done with a vague prompt (bad output) vs a structured prompt (excellent output). The difference is dramatic and immediately convincing. Use a real task from their work, not a demo.
“I don’t have time to learn another tool”
The fear: Learning curve will slow me down.
The fix: Don’t ask them to “learn AI.” Give them 5 pre-built prompts that work for their specific role. Copy, paste, fill in the brackets, submit. When they see a 2-hour task completed in 15 minutes, they will be curious enough to learn more on their own.
“What about data confidentiality?”
The fear: Uploading company data to AI is a security risk.
The fix: This is a legitimate concern, not resistance. Address it directly: use enterprise versions (ChatGPT Team/Enterprise, Claude for Business) that do not train on your data. Create a clear policy for what data can and cannot be uploaded. Involve your IT security team early so they are allies, not blockers.
“My manager isn’t using it, so why should I?”
The fear: No leadership buy-in signals this is optional.
The fix: Leadership has to visibly use AI themselves. When the CPO asks “did you run this through Claude first?” in a review meeting, the signal is clear. AI adoption is top-down in practice, even if the interest is bottom-up. If your leadership isn’t engaged, that is change management problem #1.
The 90-Day Change Management Playbook
Days 1-30: Create Visible Wins
Objective: Make AI benefits impossible to ignore with one or two showcase examples from real team tasks.
Pick 2-3 team members who are naturally curious. Give them structured prompts for their most time-consuming recurring tasks. Track time saved. In 2 weeks, have them present their results to the team informally: “This used to take me 3 hours. Now it takes 20 minutes.”
Key principle: Peer evidence beats management mandate every time.
Days 31-60: Build Habits, Not Training
Objective: Move from “I use AI sometimes” to “AI is part of how I work.”
Create a team prompt library on a shared Google Doc or Notion page. For each recurring task (spend analysis, contract review, stakeholder emails, RFP drafting), have a pre-built prompt template. Run weekly 15-minute “AI tip” sessions where someone shares a new use case. Make it conversational, not a lecture.
Key principle: Lower the friction. If using AI requires more effort than not using it, people will stop.
Days 61-90: Embed in Process
Objective: Make AI the default, not the exception.
Update your team’s standard operating procedures to include AI steps. Example: “Before presenting a category strategy, run the data through ChatGPT for anomaly detection.” Add AI-generated outputs as a standard section in review templates. When the process includes AI by default, individual motivation matters less.
Key principle: Systems beat willpower. Embed AI in the process, not in the person.
How to Measure If Adoption Is Working
| Metric | How to Track | Target at 90 Days |
|---|---|---|
| Active Users | Licence usage reports or team survey | 70%+ of team using weekly |
| Time Saved | Before/after time tracking on 3 key tasks | 30%+ reduction on tracked tasks |
| Prompt Library Usage | View count on shared prompt doc | Weekly access by 50%+ of team |
| Quality Perception | Monthly 1-question survey: “How useful is AI in your daily work?” (1-5) | Average 3.5+ out of 5 |
The Tool Is 20% of the Challenge. The People Are 80%.
Every failed AI implementation I have seen had the same root cause: great technology, poor change management. The teams that succeed invest as much in people as in tools. They create visible wins early, build habits through repetition, and embed AI in the process so it stops being optional. That is the playbook. It is not glamorous. But it works.
Need a Structured Training Programme?
Our Corporate AI Training is built specifically for procurement teams. It includes change management, prompt libraries, and hands-on workshops designed to get teams from AI-curious to AI-capable in 90 days.
Asmaa Gad is the founder of SupplyChain AI Pro, helping procurement and supply chain professionals master AI tools for real work.