Tag: AI workflow

The Board Said “So What?” After a Deck That Took 6 Hours to Build.

A client — head of strategy at a mid-sized financial services firm — came to me after what she described as “the most embarrassing board meeting of my career.” She’d used AI to build a 22-slide strategic review. The structure was immaculate. Clear sections. Logical flow. Data on every slide. The AI had done exactly what she’d asked: organise the quarterly results into a coherent deck.

She presented for eighteen minutes. The board listened politely. Then the chairman said five words that made her stomach drop: “What do you want us to do?”

She had the data. She had the structure. She had the logic. What she didn’t have was a reason for anyone in that room to care — or act. The deck was informative. It wasn’t persuasive. And in a boardroom, informative without persuasive is just a well-organised waste of everyone’s time.

When we audited the deck together, the problem was obvious. Every slide followed the same pattern: here’s what happened, here are the numbers, here’s the next slide. No context for why the numbers mattered. No connection to what the board actually cared about. No emotional stakes. The AI had produced a report disguised as a presentation.

This is the gap that nearly every AI-generated presentation falls into. Not a structure problem. A persuasion problem. And it’s a gap that AI can’t close on its own — because making AI slides persuasive requires something AI doesn’t have: knowledge of what your specific audience fears, wants, and needs to hear before they’ll say yes.

The Structure-Persuasion Gap: Why AI Output Feels Flat

AI is remarkably good at one thing: organising information logically. Give it data, a topic, and a prompt, and it will produce sections, headings, bullet points, and a sequence that makes rational sense. This is genuinely useful — it handles the tedious structural work that used to take hours.

But structure and persuasion are different skills. Structure answers “What information goes where?” Persuasion answers “Why should anyone care?” A well-structured deck can be completely unpersuasive. An unstructured but emotionally compelling argument can move a room. The ideal presentation has both — and AI consistently delivers only the first.

Here’s why. Persuasion requires three things AI doesn’t have access to: the specific context your audience is operating in, the emotional stakes attached to the decision, and the proof points that this particular group of people will find credible. AI can’t know that the CFO is worried about Q3 cash flow, that the board rejected a similar proposal six months ago, or that the CEO responds to client stories but switches off during spreadsheet reviews. These are human-intelligence inputs, and they’re exactly what transforms a structured deck into a persuasive one.

The reason most AI presentations fail isn’t that the AI is bad. It’s that the human skips the layer that makes AI slides persuasive, assuming structure is enough.

The S.E.E. Formula: Story, Evidence, Emotion

The S.E.E. formula is the persuasion layer you apply after AI has handled the structure. It stands for Story, Evidence, Emotion — three elements that, when woven into an AI-structured deck, transform it from a report into an argument that moves people to act.

Think of it this way: AI builds the skeleton. S.E.E. adds the muscle, the nervous system, and the heartbeat.

Each element serves a different persuasion function. Story provides context and makes your point memorable. Evidence provides credibility and makes your case defensible. Emotion creates urgency and makes your audience care enough to decide. A presentation that has all three is extremely difficult to dismiss. A presentation missing any one of them has a predictable failure mode.

Layer 1: Story — The Context AI Doesn’t Know

Story in a business presentation doesn’t mean “once upon a time.” It means context — the specific situation that makes your recommendation relevant, urgent, and grounded in reality.

AI output typically starts with the general: “Market conditions have shifted.” “Customer satisfaction has declined.” “Revenue targets are at risk.” These statements are accurate but they don’t anchor to anything your audience can feel. They’re abstract. And abstract doesn’t persuade.

The S.E.E. Story layer asks you to add one specific, concrete example to each major section of your deck. Not fiction — a real situation from your organisation that illustrates the point.

For example, instead of AI’s “Customer churn has increased 12% year-over-year,” the Story layer adds: “When I spoke with three of our enterprise clients last month, two mentioned they’re evaluating competitors for the first time in four years. One said — and I’m quoting directly — ‘Your platform used to be ahead. Now it’s keeping pace.’ That’s the shift the 12% represents.”

Now the board isn’t processing a number. They’re processing a threat. The data hasn’t changed. But the context makes it matter.

This is something AI fundamentally cannot generate — because it doesn’t know which clients you spoke to, what they said, or which anecdote will land with this particular audience. It’s human intelligence applied to AI structure.

Layer 2: Evidence — Turning Data Into Proof

AI is very good at including data. It’s surprisingly bad at turning data into proof. There’s a crucial difference.

Data is a number. Proof is a number plus its implication. AI will give you “NPS declined from 72 to 61.” That’s data. Proof sounds like: “NPS declined from 72 to 61 — a drop below the threshold where enterprise clients typically begin vendor reviews, based on our last three contract cycles.”

The Evidence layer in S.E.E. asks you to do three things with every data point AI generates:

First, contextualise it. What does this number mean relative to a benchmark your audience recognises? Industry average, last quarter, a target they set, a competitor’s performance. Data without context is just a number. Data with context is a signal.

Second, source it credibly. AI often presents data without attribution. Executives discount unsourced numbers. Add where the data came from — even “based on our Q3 finance review” adds credibility. If it’s external data, name the source. If it’s your own analysis, say so.

Third, connect it to consequence. What happens if this number continues? What happens if it reverses? The consequence is what transforms data from interesting to actionable. The Insight-Implication-Action framework from the course formalises this — every data point needs an insight (what it means), an implication (why it matters), and an action (what to do about it).

This evidence layer is where AI-enhanced presentations diverge from AI-generated ones. The AI handles the organisation. You handle the meaning.

Layer 3: Emotion — The Decision Trigger

This is the layer most professionals skip, and it’s the one that matters most for executive decisions.

Executives don’t make decisions based on logic alone. Research in decision science consistently shows that emotion drives action — logic justifies it afterward. A presentation that’s logically perfect but emotionally flat produces “let me think about it.” A presentation that creates the right emotional response — urgency, opportunity, risk — produces “let’s move on this.”

The Emotion layer isn’t about manipulation. It’s about connecting your recommendation to something your audience genuinely cares about. Every executive in every meeting has emotional stakes: protecting their team, delivering on promises they’ve made, avoiding the embarrassment of backing the wrong initiative, capitalising on an opportunity before a competitor does.

AI can’t identify these emotional stakes because they’re not in any dataset. They’re in the politics, relationships, and pressures of your specific organisation. Only you know that the VP of Operations is under pressure to show efficiency gains. Only you know that the CEO mentioned supply chain risk at the last all-hands meeting. Only you know that this proposal’s biggest blocker lost a similar bet two years ago and is risk-averse as a result.

The Emotion layer asks one question for each key slide: “What does my audience feel about this — and what do I need them to feel instead?” If the current state is complacency, you need urgency. If the current state is fear, you need confidence. If the current state is scepticism, you need proof that reduces perceived risk.

This is the layer that took my client’s deck from “so what?” to a follow-up meeting where the board asked her to accelerate the initiative. Same data. Same structure. Different emotional framing.

Applying S.E.E. to Any AI Deck in 20 Minutes

Here’s the practical workflow. You’ve used AI to build your deck — structure is solid, data is in place, flow makes sense. Now apply S.E.E. in three passes:

Pass 1: Story scan (5 minutes). Review each major section. For each one, ask: “Is there a specific, concrete example from our organisation that illustrates this point?” Write one sentence per section — a client conversation, an internal metric, a project outcome, a competitor move. You’re adding the anchor that makes abstract data feel real. If you can’t find a story, the section may be filler.

Your AI workflow handled the structure. This pass handles the meaning.

Pass 2: Evidence upgrade (5–10 minutes). Review every data point. For each one, add: context (vs what benchmark?), source (where did this come from?), and consequence (what happens if this continues?). Delete any data that doesn’t have a clear implication. More data with no context is worse than less data with clear meaning. Senior leaders don’t need all the information — they need the right information, framed so the conclusion is obvious.

Pass 3: Emotion check (5 minutes). For each key decision slide — recommendations, proposals, asks — answer: “What does my audience currently feel about this topic? What do I need them to feel? What one change to this slide creates that emotional shift?” Sometimes it’s reframing the opening line. Sometimes it’s adding a consequence slide. Sometimes it’s removing a defensive caveat that signals your own uncertainty.

Total time: roughly 20 minutes on top of whatever the AI took to generate the deck. That 20 minutes is the difference between “good presentation” and “approved.”

How do I make AI presentations more persuasive?

Apply the S.E.E. formula after AI handles structure: add Story (specific examples from your organisation), upgrade Evidence (contextualise every data point with benchmarks and consequences), and layer in Emotion (connect your recommendation to what your audience cares about). This 20-minute review transforms AI output from informative to actionable.

Why do AI-generated presentations feel flat?

AI excels at logical organisation but lacks access to three persuasion inputs: the specific context your audience operates in, the emotional stakes attached to the decision, and the proof points this particular group will find credible. Without these, AI produces structured reports rather than persuasive arguments.

What is the S.E.E. formula for presentations?

S.E.E. stands for Story-Evidence-Emotion. Story provides concrete, real-world context that makes abstract data feel tangible. Evidence transforms raw numbers into proof by adding benchmarks, sources, and consequences. Emotion connects your recommendation to what your audience fears, wants, or needs — the trigger that turns understanding into action.

Related reading: The presentation was perfect — the Q&A lost the deal — once your deck has the persuasion layer, prepare for the decision-making conversation that follows.

Your next step: Take the last AI-generated deck you built. Run the three S.E.E. passes: Story scan (add one concrete example per section), Evidence upgrade (contextualise every data point), Emotion check (connect each recommendation to what your audience cares about). Twenty minutes. And if you want the complete system — S.E.E. plus AVP, 132 Rule, data storytelling, and AI prompt templates for each — AI-Enhanced Presentation Mastery (£249) gives you everything in one self-study programme.