[LWN] 😔 I Lost It At The Zoo...

Lunch With Norm's Weekly Newsletter - Amazon News & Updates

😔 I Lost It At The Zoo...

What you’ll find in this week’s newsletter:

  • 😔 I Lost It At The Zoo... [True Story]

  • šŸ¾ Retail’s 2025 Year End Review

  • 🚨 Amazon’s AI Crossroads

  • āš’ļø How to Build a Launch Kit in Seconds

  • šŸ’° How I Made $1,000,000 By 18...

😔 I Lost It At The Zoo… [TRUE STORY] 

Back in the '80s, I landed my first real paying gig as a photographer for the Toronto Zoo.

It was a dream job.

I had the "Golden Ticket" to areas of the zoo the public never sees.

I felt like I owned the place.

One afternoon, I’m walking the grounds when I spotted a school group.

As it turned out, my friend Laura’s sister was in the class.

We chatted for a bit, said our hellos, and I went on my way.

A little later, I happened to pass by the baboon enclosure.

Now, you have to understand the setup.

There was a primary fence to keep the animals in, and a secondary safety fence—clearly marked with "DANGER: DO NOT CROSS".

I rounded the corner and my heart nearly stopped.

The teacher, the person in charge of these kids' safety was actually encouraging them to climb over the safety fence.

They were hanging off the primary enclosure, inches away from the mesh, trying to feed the baboons.

I didn't think, I just reacted.

I ran over screaming.

I was yelling at the teacher to get them back, yelling at the kids to move.

They looked at me like I was a madman.

It wasn't until Laura’s sister saw the sheer panic in my eyes that she realized this wasn't a joke and ushered the kids back.

When the dust settled, the teacher looked at me, completely unfazed.

She thought they were just "cute monkeys."

She had no idea that a baboon is a highly unpredictable primate that could have pulled a child's arm through that fence in a heartbeat.

She saw a "fun photo op."

I saw a disaster in the making.

Some Beardos may remember a similar story I wrote last year about a childhood friend who lost his arm due to getting to close to a ā€œcuteā€ polar bear.

Business Lesson

Don't mistake a lack of visible danger for the absence of it.

In business, the things that look like easy money or too good to be true usually have the sharpest teeth.

— Norm

šŸ‘‡AMAZON NEWS AND UPDATESšŸ‘‡

🚨 Amazon to change refund policy for Fulfilled by Merchant orders

šŸ›‘ Amazon halts plans for drone delivery in Italy 

šŸŽ„ Retail closes the holiday season leaner and more focused 

šŸš€ Amazon appoints longtime AWS exec Peter DeSantis to lead new AI org 

šŸŽ Returning gifts? Here's return policies at Amazon, Target, Walmart, more

šŸ¤– Amazon tightens platform access as AI shopping agents expand 

šŸ”„ TikTok has dethroned YouTube and Instagram and become the top social app for news among young people

šŸ¾ Retail’s 2025 Year End Revie

In 2025, retail was shaped by real pressures, tighter consumer wallets, rising costs, and the need to make digital transformation genuinely useful.

Consumers traded down where it made sense, leaned into deals, and treated credit, rewards, and instalment options as tools to manage budgets.

Across the year, two consistent themes emerged:

  • Retailers adapting how they sell, not just what they build.

  • Consumers becoming more deliberate with where they spend.

Tech Matters… But Only When Customers Don’t Notice It

A recurring message from retailers this year was simple: technology only works when customers barely notice it.

CVS leaned into what most retailers already have but often underuse, transaction and loyalty data.

At Ulta Beauty, the focus was on discovery. Ulta’s approach treats search, browsing, and buying as one fluid experience because customers want answers.

Furniture.com tackled friction differently: normalizing inconsistent product data across partners, they cut down the time people needed to compare and decide, reducing decision fatigue without removing choice.

Whatnot brought another lesson: retail can be entertainment and commerce without losing either with users spending 80 minutes a day in the app.

AI is something customers feel through better search, faster onboarding of products, smarter replenishment, and more relevant offers.

Amazon vs. Walmart

Across 2025, the biggest players in U.S. retail continued to tug on consumer dollars, even as overall spending slowed in some categories.

  • Amazon and Walmart combined generated about $1.3 trillion in retail sales over the past 12 months roughly 17% of total U.S. retail spend.

  • Amazon accounted for about 9.1% of U.S. retail sales.

  • Walmart landed around 7.6% but with a heavy skew toward grocery and everyday essentials.

Amazon Isn't Really a Retailer Anymore

Maybe the most underreported shift of the year is what Amazon didn’t say in its Q3 earnings: only 7 of its 35 bullet points focused on retail.

The rest?

Advertising, AWS, AI infrastructure, and logistics.

The numbers tell the full story…

Retail sales (online + physical stores) brought in $73 billion in Q3. But its services, third-party seller fees, ads, subscriptions, AWS, and other tools, generated $107 billion.

Retail now makes up just 40.5% of Amazon’s revenue. That’s down from 60% in 2018. And it’s not just about volume. It’s about margin.

Amazon’s ad business alone pulled in $17.7 billion last quarter, growing 22% YoY. Third-party seller fees added $42.5 billion, with Amazon taking ~50% of seller revenue via stacked fees. These services aren’t just padding the bottom line, they are the business.

Black Friday 2025

Black Friday captured the year’s consumer mindset in one snapshot:

  • 151 million shoppers made at least one purchase down 7% from the prior year.

  • In‑store shoppers dropped 11%, while online purchases increased 7%.

  • 77% of shoppers bought digitally via phone or computer.

Shoppers bought fewer items, prioritizing what mattered most without abandoning overall spend.

AI also showed up in purchase behavior: half of all shoppers relied on conversational AI tools to help complete purchases.

Nearly 60% of consumers living paycheck to paycheck used credit cards in fixed installments to manage payments.

Buy Now, Pay Later usage grew slightly to 11.9% of purchases. Black Friday increasingly looked like a financing event as much as a discount event, with direct implications for issuers, BNPL providers, acquirers, and fraud teams.

šŸ”„ Read more here

🚨 Amazon’s AI Crossroads: Fight the Bots or Power Them?

Amazon is staring down what might be the most important strategic decision in modern retail: whether to block AI shopping agents or embrace them.

Andy Jassy saw this coming.

OpenAI, Google, Microsoft, and Perplexity have all released tools that allow consumers to research, compare, and even purchase products without ever visiting a retailer’s website.

Instead of opening Amazon or Walmart, a shopper can ask a chatbot to find the best option and complete the transaction inside the AI interface.

That shift puts Amazon in a difficult position.

These agents threaten Amazon’s direct relationship with customers, its data advantage, and its margins. At the same time, ignoring them could mean losing relevance as shopping behavior changes.

Why AI Agents Are a Real Threat

Agentic commerce could drive $1 trillion in U.S. retail revenue by 2030. Morgan Stanley expects nearly half of American shoppers to use AI agents within the same timeframe, potentially adding $115 billion in e‑commerce spend.

Around 40% to 50% of Americans already use AI tools for product research. While only a small percentage currently start purchases through AI, that number is climbing quickly, especially during peak shopping periods.

If purchases move upstream into AI tools, Amazon becomes just another supplier in someone else’s workflow.

In cases like ChatGPT, the AI provider collects a transaction fee. Retailers risk paying a toll on someone else’s highway for the same transaction.

Amazon’s Defensive Posture

Amazon updated its website code to block external AI agents from crawling its site. As of this week, Amazon has blocked 47 bots, including those from every major AI company.

Amazon has also gone on offense legally.

In November, it sued Perplexity over its Comet browser agent, accusing the startup of disguising its bots to scrape Amazon’s site and make purchases without permission. Perplexity called the lawsuit a bully tactic, but the message from Amazon was unmistakable.

Rufus, Amazon’s shopping chatbot, launched earlier this year and has steadily expanded across the site.

Amazon is also testing ā€œBuy For Me,ā€ an agent that allows shoppers to purchase items from other websites without leaving Amazon’s app.

The Cracks in Agentic Commerce

OpenAI’s Instant Checkout tool, launched in September, works with only a limited set of retailers and products.

Users can only buy one item at a time and can’t link loyalty programs like Walmart+.

There are also reliability issues.

Perplexity’s Instant Buy tool encountered repeated errors while trying to purchase in-stock items. In another test, ChatGPT recommended a Breville espresso machine but displayed an image of a garden rake.

AI agents scrape and synthesize data from across the web, but retailers don’t control how that data is interpreted or presented.

For a company like Amazon, which has spent decades refining search relevance, reviews, and ranking systems, that loss of control is unacceptable.

So the big question is, will Amazon remain the platform where shopping decisions are made, or becomes the infrastructure behind someone else’s interface?

šŸš€ Read more here

šŸŒŽ Where in the World is The Beard Guy?

Can you find Norm in the picture below? Scroll to the bottom of this newsletter to see the answer!

āš’ļø How to Build a Launch Kit in Seconds

Level: Easy

1. Write a one-page creative brief in ChatGPT

We started in ChatGPT and drafted a tight brief so every tool downstream had the same direction. We included: the offer, the audience, the main promise, the tone, and a simple visual vibe (clean, bold, modern).

Then we added a small ā€œconsistency packā€: three brand adjectives, two must-have visual cues, and one hard ā€œno.ā€ This keeps you from chasing ten styles and losing half your morning.

Example prompt:

ā€˜Create a creative brief for a solo course creator launching a new offer. Include tone, audience, key message, visual vibe, 3 brand adjectives, 2 visual cues, and 1 thing to avoid. Keep it short and practical.’

2. Generate the copy blocks you’ll reuse everywhere

Next, we asked ChatGPT for production-ready copy, not vague ideas. We generated:

  • 5 headline options (short enough for a hero graphic)

  • 5 subheads (one sentence each)

  • 6 benefit bullets (plain language)

  • Captions for LinkedIn, X, and Instagram

  • A 15-second video script with on-screen text cues

3. This is where the ā€œprompt formulaā€ mindset matters. You want structured output you can paste into Canva and your video timeline, with minimal rewriting

4. Create your ā€œbase lookā€ with ChatGPT image generation

Before we went wide with image variations, we used ChatGPT to create a first visual direction we could control. We specified framing, lighting, style, and negative constraints (what not to include).

We also generated a transparent-background version for simple overlays and sticker-style elements. That made it easy to place the same motif across posts, thumbnails, and video scenes.

Example prompt:

ā€˜Create a clean, modern hero background for a course launch. Leave empty space for headline text. High contrast, studio lighting feel, minimal clutter, no logos, no tiny unreadable text.’

5. Expand into a consistent image set in Midjourney

Once the base look was clear, Midjourney helped us generate a small batch of options that still felt like the same brand. We used the brief as the core, then added composition and lighting cues, plus style references when needed.

The key move was quality control. We generated 8–12 options, picked the top 2–3, then refined them using editing actions like panning, zooming, and inpainting to fix small issues without changing the whole image. That’s how you keep consistency without endless rerolls.

6. Build a reusable template kit in Canva Magic Studio

With the visuals and copy ready, Canva became the ā€œproduction floor.ā€ We built:

  • A hero graphic layout

  • One square post template

  • One story/reel cover template

  • One thumbnail layout

7. Then we used Magic Studio tools to move fast: Background Remover and Magic Eraser for cleanup, Magic Grab and Magic Expand for quick adjustments, and Magic Write to tighten copy when it was a few characters too long.

Finally, we used Magic Switch to resize into multiple formats without redesigning each one. This is where entrepreneurs win time back, because the next launch becomes a swap-in process.

8. Turn the kit into a 15-second vertical video

We used the script from ChatGPT and created a simple storyboard: hook, promise, three quick proof points, and a clean close.

From the AI Filmmaking workflow mindset, we treated this like a mini production: a shot list of four scenes, consistent visuals across each scene, and audio that matches the pace. If you want motion, you can animate scenes (image-to-video), then add a voiceover with a tool like ElevenLabs or use text-led pacing with music.

We finished by doing one quick polish pass: timing, readability, and a final export sized for vertical. You end up with a postable video and a repeatable structure you can reuse every week.

If you would like to learn more about AI check out FUTUREPEDIA

šŸ”„ How I Made $1,000,000 By 18...

Here are my favorite tips from this episode!

1. Build funnels backward from the highest-priced offer

Rather than starting with free content or low-ticket products, the strategy begins by defining the $10K+ offer first and designing every step upstream to logically lead to it. Sellers who launch lead magnets without knowing the final sale end up with traffic that cannot convert.

The practical takeaway is to map the entire monetization ladder before building a single page or running a single ad.

2. Use paid entry offers to filter buyers

Makenna emphasizes avoiding free traffic when selling high-ticket offers. Small paid entry points ($25–$27) dramatically improve lead quality and sales efficiency by self-selecting buyers.

Sellers should introduce early friction intentionally to reduce sales team waste and improve close rates, even if that means fewer leads overall.

3. Treat webinars as math problems

Webinars in this model are run with predictable economics: lead cost, attendance rate, close rate, and revenue per attendee are all known in advance. This allows confident ad spend and repeatability.

Stop thinking of webinars as educational sessions and start treating them as repeatable conversion systems that must pass a spreadsheet test before launch.

4. Monetize the thank-you page instead of wasting it

Makenna calls out thank-you pages as the most underutilized asset in marketing, yet they represent the moment of highest attention. Sellers who simply say ā€œcheck your emailā€ are discarding paid traffic value.

Every thank-you page should include a next action, upsell, or pre-framing step that moves the buyer further into the funnel.

5. Replace ā€œbook a callā€ with application funnels for high-ticket sales

Direct calendar booking produces low-quality conversations and poor show-up rates. But funnels increase commitment, gather sales intelligence, and frame the call before it happens.

Sellers offering anything above $1K should gate calls with a brief application to improve both conversion and experience.

šŸ”„ Watch the full episode here

Find this episode of Marketing Misfits on YouTube and anywhere you listen to podcasts

šŸŒŽ MARK DOWN THESE EVENTS! šŸŒŽ

Feb 10-12 - Dubai - Free

Feb 23-26 - Palm Springs - $499.50

Mar 9-11 - Las Vegas - $895 (For Non Members)

April 21 - 23 - Fort Lauderdale - $799

Mar 10-12 - Las Vegas - Starting at $99

→ FULL CALENDAR ←

And that’s it, Beardos.

See you next Monday!

- Norm

Interested in sponsoring the Lunch With Norm Newsletter? Reply back to this email for more details.

Reply

or to participate.