Why 90% of AI Presentations Fail (And How to Fix Them)

During my consulting days, I spent countless hours perfecting PowerPoint slides. It was almost comical – your rise through the consulting ranks seemed directly tied to your slide-making prowess.

Eventually, I realized this wasn't a skill I wanted to master forever. Life's too short for endless formatting tweaks!

Then 2022-2023 happened. Suddenly AI tools were everywhere, all promising the same thing: "Create stunning presentations in minutes!"

(I've shared quite a few posts about these tools myself...)

Gamma, ChatGPT-o3, MS Copilot, Claude, Gemini – they all can indeed produce visually impressive decks in minutes.

But, they always…sound average: feed them shallow insights, and you'll still face that awkward silence when your audience realizes there's no substance behind the pretty slides.

Today's newsletter reveals the three-step workflow that transforms AI from mere slide decorator into deal-closing, impactful presentation powerhouse.

Why AI Presentation Tools Stumble

Let’s be honest.

AI tools like Gamma are great for low-stakes stuff —

Team updates, product overviews, internal briefs.

But in high-stakes business scenarios —

Client pitches, investor decks, market entry reports —

They fall short.

Not in how they design the slides, but in what goes on them.

Why?

Because they don’t generate insights.

They present them.

There’s a gap between insight generation and insight presentation.

And you need both.

AI shines at presentation, persuasion eventually, but not depth. Most tools live in the “design” layer, so they’ll happily polish fluff if you skip real insight generation. Think baking: ChatGPT is the sous-chef, Gamma is the decorator—hand them boxed mix and you still serve boxed mix.

So here’s what I’ll show you in this newsletter:

• How I use ChatGPT + Gamma to recreate a real pitch deck I once delivered
• The 3-step system: Research → Structure → Design
• Real examples, slide edits, and pro tips
• What AI can’t do (yet), and how you step in

Let’s dive in.

The 3-Step “Research → Structure → Design” System

STEP 1 – Generate Real Insight

Prompt

Objective: Provide a market-entry report for a sport nutrition brand targeting men 25-45
Markets: France, UK, Germany
Framework: Porter’s Five Forces
Deliverables: Exec summary • ranked scorecard • tactical recs

Key findings (June 2025 data)

• Attractiveness ranking (Germany > France > UK): Germany’s sports-nutrition sales are projected to rise at ≈5.3 % CAGR to 2030, under a relatively open NemV regime; France’s market is ~US $526 m with 4.8 % CAGR potential; the UK is larger (US $1.08 bn) but scores lowest on rivalry and buyer power. [Source1] [Source2] [Source3]
• UK scale vs. pressure: The UK leads in value and growth (8.7 % CAGR) but faces intense D2C competition—Myprotein alone turned ~£580 m in 2024—and consumers remain cost-sensitive despite a record 11.5 m gym memberships. [Source1] [Source2] [Source3]
• Supplier cost squeeze: European whey-protein isolate prices jumped about 13 % YoY in 2024, compressing margins—an issue amplified for import-dependent France and highly price-competitive UK brands. [Source1]
• Regulatory nuances: Germany’s NemV largely mirrors EU provisions, lowering entry hurdles; France’s ANSES continues to flag health-risk vigilance on sports supplements; post-Brexit UK guidance demands dual GB/NI labelling and strict supplement claims. [Source1] [Source2]
• Substitution threat accelerating: Europe’s plant-based protein market is on track for US $18.6 bn by 2030, and functional-beverage sales should hit US $45.6 bn, eroding pure-powder share and rewarding diversified product lines. [Source1] [Source2]

Pro-tip: Use ChatGPT’s Deep Research mode (raised to 25 queries/month for Plus & Team plans) so you’re not throttled mid-project openai.com.

STEP 2 – Shape the Story

Toggle Deep Research off and instruct ChatGPT:

Create an 8-slide outline that convinces the CMO to pick Germany
Include one slide explaining why not France/UK; use SCQA.”

⚠️ A critical point: a presentation is about conviction, not just slides. Use Step 1 to form your opinion and craft a narrative arc angles, then let AI help shape that story. This is where your true value lies - and exactly what separates an "insightful" presentation from an "average AI-generated" one.

Outline highlights:

1. Situation: Europe’s sports-nutrition market is on a tear—US $12.3 bn (2023) on course for US $25.5 bn by 2030
2. Complication: France, UK and Germany all show healthy mid-single- to high-single-digit CAGRs (4.8 %–8.7 %) but differ sharply in competition, regulation and input costs
3. Question: Where will €1 spent earn the highest NPV?
4. Answer: Germany wins — 5.3 % forecast CAGR, low market concentration and consumer spend rebounded to 11.7 m gym members in 2024
5. Germany: Demand & growth upside
6. Germany: Competition & channel headroom
7. Germany: Supply & regulatory clarity
8. Why not France or UK?

This narrative structure is very interesting - I've rarely seen such clear storytelling even from consultants with years of experience. Trust me, once you start using AI to brainstorm your presentation narratives, you'll wonder how you ever did without it.

STEP 3 – Make It Beautiful (Fast)

1. Copy outline → Gamma → “Create with AI → Presentation.”

2. Settings I swear by:

1. Text: Condense (outline is already solid).
2. Images: Illustrations first, swap to Unsplash later.
3. Theme: Icebreaker – Light Blue (boardroom-friendly).

3. Gamma spits out an 80 %-ready deck in ~60 seconds.

Human TLC (the final 20 %)

• Rewrite titles with insight: “Capturing Indonesia’s 54 % YoY Acne-Wash Surge.”
• Replace odd AI art with German specific shots.
• Add a comparison table—Gamma’s “Add Card → Table” keeps formatting.
• Use Spotlight Mode (S during a presentation) to blur future content and keep eyes where you want them.
• Remove the tiny Gamma watermark: Edit Theme → toggle off [How-To Blog].

Speed Hacks & Gotchas

Beyond the Boardroom

• Career-pivot deck: weigh industries—Porter for market size, GPT for skill gaps, Gamma for visuals.
• Non-profit pitch: swap Porter’s forces for impact metrics.
• Wedding slideshow: research venues, structure costs, let Gamma add romance (and spare your sanity).

Key Takeaways (stick these on your monitor)

1. Insight > Aesthetics. Research first; pretty follows.
2. Right tool, right time. Research → Structure → Design.
3. Human–AI collaboration beats either alone. As Wharton’s Ethan Mollick says, the magic happens at the intersection—we choose the collaboration points.

That’s it my friends. Now you have it; sauce to creating killer presentations with AI is now in your hands.

Remember the winning formula: solid research + your unique perspective and narrative angles + ChatGPT for storyline structuring + Gamma with just the right settings for a nice visual draft + the remaining 20% are on you.

If you've been following my letters for a while, you'll notice a familiar theme: AI's true power isn't about finding that one magical tool that does everything. It's about weaving multiple tools together into powerful workflows that amplify your capabilities.

Try this with your next presentation this week. I'd love to hear how it goes!

All the best,

— Charafeddine