35 minutes • Evaluate team capabilities and design targeted learning paths
Organizations often make the mistake of purchasing training programs before understanding what their team actually needs. They might invest in advanced AI prompt engineering training when their team's real need is basic familiarity with language models. Or they might offer comprehensive training to people who already have the knowledge, wasting time and resources.
Effective training starts with assessment: understanding where your team stands, what gaps exist, what individuals need, and what learning pathways will actually solve problems. Assessment reveals that "AI literacy" isn't monolithic—different roles need different skills. Researchers need different AI knowledge than writers or operations staff. Assessment ensures training is targeted, relevant, and valuable.
Team members typically fall along a spectrum: some have no AI experience and need foundational education; some have dabbled with ChatGPT but don't understand how to use it strategically; some use AI tools regularly but without deep understanding of capabilities and limitations; some are advanced users pushing tools to their limits. Effective training acknowledges this spectrum and doesn't try to treat everyone identically.
Create a simple assessment: What AI tools have you used? For how long? What tasks do you use them for? How confident are you using AI? Rate your skills (beginner, intermediate, advanced). Use this to build a profile of your team's current state.
Beyond general AI literacy, assess knowledge of specific tools your organization uses. Does everyone know how to access and use Claude? Have they used Airtable? Do they understand your automation systems? Tool-specific knowledge determines whether people can actually use available resources.
Create a tool knowledge matrix: rows are team members, columns are tools. Mark whether each person is unfamiliar, somewhat familiar, or proficient with each tool. Identify gaps. These gaps are training priorities.
Competence and confidence are different. Someone might be technically capable but lack confidence, so they don't use tools effectively. Someone else might be overconfident about their abilities. Assess both capability and confidence. Ask: "How confident are you using this tool effectively?" Confidence gaps might be addressed through mentoring or reassurance; capability gaps require education.
Brief one-on-one interviews reveal more than surveys. Thirty-minute conversations help you understand each person's actual needs, concerns, learning preferences, and barriers. You learn that someone avoids AI tools because they've had bad experiences, not because they lack ability. You discover that someone is more capable than their survey indicated. Interviews build relationships while gathering information.
Once you know current state, identify gaps: what skills do people need to succeed in your organization's AI-enabled future? If your strategy is using Claude for proposal drafting, writers need Claude skills. If you're automating grant discovery, researchers need to understand how the automation works and how to interpret results. Not all skills are equally critical.
Create a critical skills list for each role: what must this role be able to do? Writers need to write with AI assistance. Researchers need to evaluate AI-researched information. Administrators need to understand systems without necessarily operating them. Map gaps between current capabilities and critical skills.
Some skill gaps significantly impact organizational performance; others are nice-to-have. Prioritize high-impact gaps: skills that, when learned, enable people to contribute more effectively. If learning prompt engineering would improve proposal quality dramatically, it's high-priority. If learning advanced data visualization is interesting but not critical, it's lower priority.
Also consider urgency. If you're implementing automation next month, automation skills are immediately critical. If you're planning AI integration six months from now, the timeline is less urgent. Let impact and urgency guide priorities.
Different roles need different training. Grant writers need different AI knowledge than researchers, who need different knowledge than administrators. Create role-based learning paths: what should each role master? What are nice-to-have skills? What's beyond their role's scope?
Example paths: Writer path includes prompt engineering, feedback interpretation, quality assurance with AI. Researcher path includes finding and evaluating AI-researched sources, combining AI output with authoritative sources. Administrator path includes system management, workflow troubleshooting, data consistency.
Within roles, people are individuals with different aspirations. Some writers want to become AI experts; others want minimal AI involvement. Some researchers love technology and want to explore advanced tools; others prefer traditional research. Individual learning paths acknowledge these differences.
Conduct career conversations: where do you want to grow? What aspects of AI interest you? What are your learning preferences? Use these conversations to customize learning paths. People invest more effort in learning toward goals they care about.
Document the competencies you need: basic AI literacy, Claude proficiency, prompt engineering, data management, automation understanding, critical thinking about AI outputs. For each person, map current competency level and target competency. This becomes the basis for individualized training plans.
Some barriers are psychological. Fear of obsolescence ("will AI replace me?"). Skepticism ("this won't really help us"). Resistance to change ("we've always done it this way"). Understanding these barriers is crucial. Training doesn't overcome them directly—change management and communication do. Understanding the barriers helps you address them.
Team members are busy. Grant work is urgent and time-consuming. Finding time for training is hard. Lack of access to tools makes learning difficult. Small budgets limit training options. Understand these constraints. Design training that works within them. Short, focused sessions beat long seminars. Peer-to-peer learning beats expensive consultants. Hands-on practice beats lectures.
People learn differently. Some are visual learners who need videos and diagrams. Others are hands-on learners who need to try tools immediately. Some learn well in groups; others prefer individual instruction. Effective training accommodates different styles: video tutorials for visual learners, hands-on labs for kinesthetic learners, written guides for readers, discussion groups for social learners.
Quick surveys assess baseline knowledge: "Have you used ChatGPT?" "How often?" "For what tasks?" Surveys are fast and can reach everyone. They're impersonal but efficient for gathering broad information. Use surveys to identify extreme cases (people with high AI experience, people with none) who might need special attention.
Ask people to demonstrate skills: write a proposal section using Claude, research a topic, use your automation system. Actual performance reveals more than self-assessment. Some people overestimate their abilities; others underestimate. Demonstration-based assessment is accurate but time-consuming.
Structured conversations reveal depth. Ask about learning preferences, concerns, goals, previous training experience. Focus groups explore common themes across people. Interviews and focus groups take time but provide rich insight that surveys miss.
A nonprofit assessed their team and found: 60% had minimal AI experience, 30% used ChatGPT casually, 10% were already using AI daily in their work. They discovered that experienced users actually wanted advanced training, while basic users needed foundational education. They also learned that some team members were eager to learn while others were skeptical. Assessment revealed these patterns, enabling targeted training rather than one-size-fits-all approach.
Assessment results guide training priorities. What are the biggest skill gaps? What barriers need addressing? What learning preferences are most common? Use this information to build a roadmap: what will you train, in what order, using what methods? The roadmap ensures training addresses real needs rather than generic options.
Share assessment findings with leadership. Explain gaps and priorities. Make the case for training investment. Data-backed recommendations are more compelling than opinions. "70% of our team lacks prompt engineering skills, which limits proposal quality" is stronger than "we should train people on AI."
Assessment results inform learning goals. Based on identified gaps, what should people be able to do after training? Writers should be able to write compelling proposals with AI assistance. Researchers should be able to evaluate AI-researched information. Administrators should understand system workflows. Specific, measurable outcomes guide training design and assessment.
Assessment isn't one-time. Skills develop. New challenges emerge. Reassess periodically: are people maintaining skills? Are new gaps emerging? Is training actually improving capabilities? Use ongoing assessment data to adjust training continuously. Learning organizations treat assessment as ongoing improvement, not one-time event.
Next, we'll explore designing effective AI training curriculum based on assessment findings.
Continue to Next Lesson