Mad Hudson Growth Playbook
Mad Hudson has three growth levers that aren't fully connected yet. Brad's 6.2M TikTok following is the biggest distribution asset in the business — and the brand doesn't own it. AI shopping agents are already deciding which watches to recommend to buyers — and Mad Hudson's product data is halfway ready. Every Whatnot stream produces content that disappears instead of compounding. This playbook connects those three things into one system.
The Brad Question
Brad Podray has 6.2 million TikTok followers and 321.9 million likes through his Scumbag Dad persona. He actively promotes Mad Hudson on his personal channels. His reach dwarfs the brand's own accounts by orders of magnitude.
This is the most important structural fact about the business. It's also the biggest risk.
Instagram (@madhudsonwatches), Whatnot streams, TikTok Shop listing, email list, website. These survive if Brad pivots to something else.
Brad's TikTok (6.2M), Brad's YouTube (~600K), Brad's Threads. Massive reach, but it belongs to Brad, not Mad Hudson.
Creator-led brands that outlast their creators are the ones that build community around the product, not just the person. Chamberlain Coffee stalled when Emma stepped back from YouTube. Feastables still depends on MrBeast. Mad Hudson's advantage: the watches have their own identity. The product can stand alone. But only if the brand builds its own audience alongside Brad's.
The content flywheel
Right now, each channel operates mostly in isolation. Connected, they compound:
The weakest link right now: Whatnot stream → email capture. A viewer who watches but doesn't buy has no path back to Mad Hudson unless you have their email.
Selling to AI Agents
When a buyer opens ChatGPT or Perplexity and types "find me a good artist collaboration watch under $250," the AI doesn't browse your website like a human does. It doesn't care about the speakeasy aesthetic. It reads structured data, clips a recommendation, and moves on.
If your product data isn't formatted for machines, the AI skips you entirely. The buyer never sees your name.
Story, emotion, identity. The feeling of discovering something real. The artist narrative. The smashed glass that's actually beautiful.
Product name, price, materials, availability, creator name. In structured, extractable format. No poetry — just facts the algorithm can clip and recommend.
You have to serve both. This is the duality.
Extractable chunks
A self-contained 1–3 sentence statement an AI can lift from your product page and use as a recommendation without any additional context.
Weak: "We think time is art. Our watches celebrate that."
Strong: "The Broken Time watch is a $240 limited-edition timepiece where the dial is deliberately fractured — the damage is the design, not a defect. Each piece is unique because no two breaks are identical."
The strong version tells the AI: price, product name, defining feature, price justification, and scarcity signal. It answers "what is this?" completely. Every flagship product needs one of these in its opening paragraph.
Where Mad Hudson stands with AI commerce
| Platform | Status | Next step |
|---|---|---|
| Google Merchant Center | 93/93 approved | Confirm auto-sync is active |
| Google AI Overviews | Schema live | Write extractable chunks |
| TikTok Shop | Broken Time only | Expand to full catalog |
| Perplexity Shopping | Unknown | Test it — search for your watches |
| ChatGPT recommendations | Unknown | Submit sitemap, structured data helps |
AI shopping agents filter out products without UPC barcodes as potential counterfeits. Mad Hudson doesn't use standard UPCs because these are limited artist runs. The fix: apply for a GS1 company prefix (~$250/year). Gives you legitimate barcodes for every product. Permanently solves the problem instead of writing custom exemption rules that break whenever the algorithm updates.
Social Media: What We Know and Don't
| Platform | Status | Data gap |
|---|---|---|
| Instagram @madhudsonwatches | Active, product reels | Follower count unknown |
| TikTok @brad_podray | 6.2M followers | Not a brand account |
| TikTok Shop | Broken Time listed | Sales data unknown |
| Page exists | Metrics unknown, low priority | |
| YouTube | No brand channel | Biggest content gap |
| Whatnot | Active | 87 viewers (Coop's stream) |
This is a 30-minute fix. Log into Meta Business Suite, pull the numbers. Everything else in content strategy is built on top of this data. Get it this week.
Content by platform
Instagram — 3–4 posts/week. Two product or aesthetic posts, one artist or behind-the-scenes, one stream announcement or recap. The speakeasy philosophy works naturally here — behind-the-scenes content feels exclusive, not promotional.
TikTok — minimum viable play: repurpose Whatnot stream clips (30–60 seconds each) the day after each stream. Zero production cost. Post on Brad's account or a brand account — either works, both is better.
YouTube — create the channel, claim the handle, publish one video. YouTube content has a long shelf life. A 3-minute "The Story of the Broken Time Watch" video gets found two years from now by someone who never heard of Brad Podray. Start with 4–6 videos per year.
The repurposing workflow
One Whatnot stream turns into five pieces of content across four channels:
The editing is the constraint. CapCut on a phone handles it. 30–60 minutes per stream.
The Moves
Ten actions, ranked by impact. The first four can happen this week.
Risks and How We Handle Them
| Risk | How we handle it |
|---|---|
| Brad deprioritizes watches | Build brand-owned audience now (email list, Instagram, YouTube). The product identity can stand alone — the watches are the thing, not the personality. |
| AI agents skip Mad Hudson | Extractable chunks + GS1 barcodes + Merchant Center feed. Make the data so clean the algorithm can't miss you. |
| Content production bottleneck | Whatnot streams are the content factory. Clip and repost. Don't create from scratch — repurpose what already exists. |
| Instagram reach unknown | Step 1 fixes this. Can't strategize on channels you can't measure. |
| TikTok platform risk | Diversify: YouTube (long shelf life), email (owned), website (SEO). Never depend on one algorithm. |
Press coverage to date
| Outlet | Coverage |
|---|---|
| The Awesomer | "Broken Time Watch Celebrates Imperfection in an Imperfect World" |
| Noissue Blog | September 2025 feature on packaging partnership |
| StayHappening | "Enter The Madness" event listing (August 2025, Commonspace Collective) |
Limited traditional press so far. The Brad video + Whatnot success story make a compelling pitch to Block Club Chicago, Worn & Wound, and Modern Retail. Each placement makes the next easier.
What's already built
| Asset | Status |
|---|---|
| Google Merchant Center | 93/93 products approved |
| JSON-LD schema markup | Live on all product pages |
| Legacy code cleanup (112KB removed) | Complete |
| SEO meta descriptions (13 pages) | Drafted, being pasted into Shopify |
| Whatnot strategy page | Live |
| GSC keyword tracking | Active since March 6 |
| Blog article draft | Ready for Connor's review |
| Live Drops page prototype | Prompt ready, needs stream schedule |
AI commerce: the full picture
Traditional SEO optimizes for a human clicking through ten search results. AEO (Answer Engine Optimization) optimizes for an AI agent reading your product data, deciding you're credible, and recommending you to a buyer who never sees a search results page.
The shift is already happening. By 2026, a meaningful percentage of product discovery starts in ChatGPT, Perplexity, or Google's AI Overview — not the traditional blue links.
What "agentic storefront" means in practice:
Metafields populated: artist name, edition size, materials, production method, collaboration type. Not visible on the page, but readable by AI crawlers.
Descriptions as structured answers: what it is, what makes it different, who made it, how to order. Predictable format the AI can parse.
Related products tagged explicitly: "if you liked CD Watch, you'll want Broken Time" should be in the data, not just implied by page layout.
Creator-led brand comparison cases
Feastables (MrBeast): Built on YouTube audience. Expanding to retail + licensing. Still depends on Jimmy Donaldson's continued relevance. Product quality keeps it alive, but discovery is personality-driven.
Chamberlain Coffee (Emma Chamberlain): Emma's step back from YouTube in 2022–23 caused visible stagnation. They're recovering by building independent brand identity. Cautionary tale for any creator-dependent brand.
PRIME (Logan Paul + KSI): Pure creator-audience play. High velocity, zero brand identity if either steps back. Volume-driven, not story-driven. Opposite of Mad Hudson's model.
Mad Hudson's advantage: the watches have their own identity. Broken Time doesn't need Brad to be interesting. But the audience that finds it currently depends on Brad's reach.
Mad Hudson Watches · Growth Playbook · March 22, 2026 · Prepared by Garen Hudson