Module 5 reference toolkit. Everything you need to run a readiness assessment — interview questions, risk mapping, and task exposure in one place.
Anne’s actual questions from 6 Meridian client engagements. A quick-reference set for every interview, plus the full annotated library by category.
Use the quick set in every interview · Pull from the full bank as needed · Adapt by role and what you already know
The Quick Reference is sequenced to build trust before going deep — ask all of Section 1 in every interview. The Full Question Bank is a toolkit, not a script. Pick by category based on who you are talking to and what you have already learned.
Sequenced to open up, not close down. Section 1 in every interview. Section 2 when you need more depth.
Ask Section 1 in every interview. These are sequenced to build trust before going deep. Section 2 has additional questions for larger organizations, more complex workflows, or when data handling is a central concern.
These are the questions from the Meridian engagement. Ask all of these. They are sequenced to open up, not close down.
Add these when you need more depth. Pick the ones that fit for this interviewee.
Anne’s annotated library from the Meridian client engagement. Every handoff you trace is a potential bottleneck.
Build rapport before anything else. Never open cold.
The move: Anne never opens cold. She matches the other person’s energy. She shares something from her own week first — sick kid, missed meeting, being behind on work. This is not small talk. It licenses the interviewee to open up.
Ask everyone. Word it differently every time.
Why different wording matters: “What do you do” gets a job title. “What do you really do in a given day” gets the truth. “What does a client engagement look like for you” traces the full lifecycle.
Walk them through their workflow step by step. You are building the process map as you go.
The skill: Follow the document, not the person. Trace what happens from the moment a request comes in to the moment the deliverable goes out. Every handoff is a potential bottleneck.
These go after friction. Give people permission to complain by calling out the emotion first.
“Double black diamonds” and “magic wand” are Anne’s go-to frames. They work because they name the emotional load. People will describe their pain points openly when you give them a safe word for frustration.
Never ask this first. It comes after trust is built and pain points are on the table.
The underlying question: Not “do you use AI” but “what is your relationship with AI.” Fear? Curiosity? Already using it secretly? Hostile? The answer tells you where to start the training. Anne never asks “are you worried AI will replace you.” She asks about behavior, not feelings.
Why this matters: Leadership will tell you where the bodies are buried. They will say things like “you have to talk to John because he is anti-AI and we want him to understand it’s time to get on board.” This question gives them permission to direct your attention.
An open door, not an assignment. The interviewee leaves feeling heard, not tasked.
16 assessment sub-tasks. For each one: what AI can do, what can go wrong, and how to protect the client.
Use this to scope your assessment · Build time savings estimates · Identify mitigation strategies before you walk in
Map each task to your specific client. Not every client does every task. Focus on the ones with the highest time savings that match what you heard in discovery. Every mitigation on this list has the same foundation: human review before anything goes to a stakeholder.
AI chatbots handle scheduling, reminders, and availability optimization.
Scheduling conflicts; experience feels impersonal to stakeholders.
AI suggests times — humans confirm all final bookings.
AI-assisted live transcription and real-time question prompts reduce note-taking burden.
Stakeholder concerns about AI recording their responses.
Explicit transparency on data use; offer opt-out before every session.
AI auto-transcription with key theme identification — the highest-impact task in the assessment.
AI summaries may misinterpret nuance or tone in sensitive conversations.
Human review of every summary before it leaves the consultant’s hands.
AI-powered surveys and analysis tools cut manual analysis in half.
Data privacy concerns from stakeholders sharing sensitive organizational information.
Establish clear anonymized data handling policies upfront.
AI-driven industry trend analysis and benchmarking delivers insights far faster.
Reliability of AI-sourced data is uneven — hallucination risk on niche topics.
Validate every AI-generated finding with at least one additional human research source.
AI scans reports, whitepapers, and published data at scale.
AI may misclassify competitor strategy or miss context-dependent signals.
Human cross-validation on all competitive claims before client delivery.
AI clustering of behavior and engagement levels surfaces patterns humans miss.
Bias in AI models can produce exclusionary or inaccurate segment definitions.
Test AI recommendations against historical data before applying to strategy.
AI-generated projections and scenario modeling speed up strategic planning.
Over-reliance on AI models without human judgment in the loop.
Use AI as a supporting tool — never the sole decision-maker on projections.
AI-driven analysis of qualitative and quantitative data cuts writing time significantly.
AI may misinterpret nuanced input from complex, multi-stakeholder environments.
Human validation of all AI recommendations before finalization.
Real-time AI tracking of engagement generates reports faster and on demand.
Risk of misreporting performance indicators if AI models aren’t calibrated.
Align AI models with the organization’s historical success metrics before use.
AI-assisted drafting based on RFPs and templates accelerates first drafts.
AI text often lacks the personalization that high-stakes proposals require.
Consultant review and edits required on every proposal before it goes out.
AI-generated summaries with key takeaways cut preparation time dramatically.
Risk of missing stakeholder-specific context that matters to board audiences.
Human review before finalization — board communications carry high stakes.
Automated approvals, meeting notes, and reminders reduce admin burden across the team.
AI errors in task prioritization can cascade into missed deadlines.
Human validation of all critical workflow steps — automation is support, not a replacement.
AI-generated summaries and task tracking dramatically reduce post-meeting work.
AI may misinterpret or miss nuanced action items, especially those tied to implicit commitments.
Consultant reviews AI notes before distributing to participants.
AI auto-formatting and structured content suggestions speed up template production.
Rigid templates may not allow enough customization for complex client needs.
Human oversight on final approval; templates are starting points, not endpoints.
AI-driven customization of communications reduces drafting time while maintaining relevance.
Over-personalization using behavioral data can feel intrusive or surveillance-adjacent.
Apply ethical AI guidelines for outreach; let humans set the personalization ceiling.
Transcription and summarization — 70–90% time savings
First drafts: reports, proposals, emails — 30–60%
Data analysis, segmentation, trend identification — 50–70%
Scheduling, reminders, workflow automation — 30–75%
Build trust with stakeholders — this is human work, zero shortcuts
Make strategic decisions — AI informs, humans decide
Navigate organizational politics — no model for this
Replace the expert-in-the-loop for quality control
The one rule that covers every sub-task on this list: Human review before anything goes to a stakeholder. Every time. No exceptions.
33 business tasks mapped to AI exposure levels E0–E6, with time savings estimates for each.
Use this to classify your client’s tasks · Focus on E1 tasks with 50%+ savings · E0 tasks stay human
Map your client’s tasks against this matrix — not every client does every task. Focus on E1 tasks with 50%+ time savings first. Flag E0 explicitly: clients need to hear that AI is not replacing their leadership. The E6 tasks are conversation starters, not deliverables. Every task above E0 carries the same requirement: human review before it goes to a stakeholder.
AI cannot meaningfully help with this task
Current LLMs (ChatGPT, Claude, etc.) can directly assist
Tools built on top of LLMs can assist — specialized SaaS, automations
Image and media generation tools add capability
Advanced conversational AI and sentiment analysis
Requires reasoning beyond current LLMs — emerging territory
These tasks require human leadership and interpersonal judgment
| Task | Time Saved | Why AI Can’t Help |
|---|---|---|
| Team Management | 0% | Requires interpersonal communication and leadership AI cannot replicate |
Quick wins with ChatGPT, Claude, Gemini, Perplexity
| Task | Time Saved | How AI Helps | Key Risk |
|---|---|---|---|
| Website Optimization | 20–40% | AI suggests improvements to copy, navigation, CTAs | May not account for brand-specific context |
| Competitive Analysis | 40–60% | Quickly gather and summarize competitor products, pricing, strategies | AI may misclassify competitor strategy |
| A/B Testing | 30–50% | Design tests, analyze results, recommend winning variations | Over-reliance on AI interpretation of statistical significance |
| Internal Communications | 50–70% | Draft memos, newsletters, announcements | May lose organizational tone |
| Crisis Management | 20–30% | Draft statements, responses, communication plans | Sensitivity and timing require human judgment |
| Data Privacy and Compliance | 20–40% | Research regulations, ensure compliance, draft privacy policies | Legal accuracy requires human verification |
| Budgeting | 50–70% | Generate budget templates, allocate resources, track spending | AI may not understand organizational constraints |
| Sales Forecasting | 40–60% | Analyze historical data, trends, seasonality | Models need human calibration |
| Performance Analysis and Reporting | 60–80% | Collect, analyze campaign data, generate reports | Risk of misreporting indicators |
| Email Marketing | 50–70% | Write copy, personalize subject lines, segment audiences | Risk of formulaic output |
| SEO Optimization | 20–40% | Suggest keywords, optimize content, analyze traffic | Rapidly changing algorithms |
Combine AI writing with purpose-built SaaS platforms
| Task | Time Saved | How AI Helps | Key Risk |
|---|---|---|---|
| Public Relations | 30–50% | Draft press releases, pitches, talking points; E2 adds PR management tools | Media relationships require human judgment |
| Event Planning | 20–40% | Concept ideation, promotional materials, follow-up comms; E2 adds event management software | Logistics still require human coordination |
| Partnerships and Influencer Marketing | 20–30% | Outreach emails, contract drafting, campaign tracking; E2 adds influencer platforms | Relationship quality depends on human interaction |
| Marketing Automation | 40–60% | Automated workflows for nurturing, email, social; E2 adds personalized campaign platforms | Over-automation can feel impersonal |
| Market Research | 30–50% | Gather and summarize data, competitor analysis; E2 adds specialized research tools | Reliability of AI-sourced data |
| Strategic Planning | 20–40% | Brainstorm strategies, SWOT analysis, predict outcomes; E2 adds advanced planning tools | Strategy requires human judgment on trade-offs |
| Campaign Implementation | 20–30% | Schedule campaigns, coordinate tasks, generate reports; E2 adds campaign management software | AI errors in prioritization |
| Lead Generation and Nurturing | 30–50% | Write lead magnets, drip campaigns, lead scoring; E2 adds lead management platforms | Qualification still needs human touch |
Text AI plus image, video, and design generation tools
| Task | Time Saved | How AI Helps | Key Risk |
|---|---|---|---|
| Brand Management | 20–40% | Maintain voice and consistency, generate creative assets, monitor mentions | Brand nuance hard for AI to capture |
| Product Launch | 30–50% | Launch strategies, promotional materials, product descriptions | AI may miss market positioning subtleties |
| Content Creation | 40–60% | Copy, social posts, blog articles, video scripts, visuals | Risk of generic output without human editing |
| Social Media Management | 40–60% | Schedule posts, draft content, analyze performance, generate visuals | Over-reliance reduces authenticity |
Sentiment analysis and personalized interactions at scale
| Task | Time Saved | How AI Helps | Key Risk |
|---|---|---|---|
| Customer Relations | 30–50% | Draft emails, handle inquiries, analyze feedback; E5 adds personalized interactions and sentiment analysis | Sensitive situations require human empathy |
Beyond current LLMs — conversation starters, not deliverables
| Task | Time Saved | How AI Helps | Why It’s Hard |
|---|---|---|---|
| Long-term Strategic Planning | 30–50% | Model scenarios, assess risks, optimize multi-year strategies | Requires anticipating market shifts with incomplete data |
| Predictive Analytics and Modeling | 40–60% | Refine statistical models, uncover hidden patterns | Complex causal relationships, not just correlations |
| Attribution Modeling | 20–40% | Assess impact of marketing touchpoints across the full client path | Deep understanding of causal relationships needed |
| Complex Decision Making | 10–20% | Weigh options with numerous variables and uncertain outcomes | High-stakes decisions require organizational context |
| Creative Problem Solving | 10–30% | Generate ideas, brainstorm solutions, evaluate feasibility | True novelty vs. recombination of existing patterns |
| Negotiation and Deal-Making | 10–20% | Analyze terms, identify negotiating positions, suggest strategies | Human dynamics, power, and relationship history |
| Adapting to Emerging Trends | 20–40% | Identify new technologies and behaviors, predict impact | Distinguishing signal from noise in real time |
| Resource Allocation Optimization | 20–40% | Multi-factor optimization across channels and constraints | Organizational politics and priorities shift |
0–30% savings — AI assists but the human does the heavy lifting.
30–50% savings — AI does the first draft; human edits and finalizes.
50–80% savings — AI does most of the work; human reviews and approves.
80–90% savings — AI handles it; human spot-checks. Nothing hits 100%. The human is always in the loop.