Current mission:
Get this to $300K ARR in the next 6 days.
No magic plan.
Just posting more videos, launching 2 new accounts, and flooding the algorithm until the number moves.
Also think we finally cracked native talking-head content with synthetic creators.
That could be the unlock.
Will not rest until it hits.
Stay tuned.
I know a 27 year old running a Pinterest affiliate operation that clears $312,000/mo & 99% of the women clicking her links think she's a stay at home mom in Vermont sharing kitchen finds
She has never been to Vermont
She is not a stay at home mom
She is a kid named Aleksandar in Belgrade
Here's the entire machine & it's the dumbest free money loop on the internet right now:
Pinterest has the lowest fraud detection of any social platform. The algorithm rewards aesthetic consistency over account age. A brand new account with 80 perfectly composed pins will outperform a 5 year old account with 4,000 mediocre pins. Pinterest does not care if you're real. Pinterest cares if your grid converts
Step 1: he scraped the top 2,400 highest-saved kitchen pins from the last 18 months. Fed them into Claude. Asked for the 14 visual patterns that print saves. Output: warm wood, beige tones, white ceramics, one ingredient mid-frame, soft window light
Step 2: he built one AI "mom" character. Mid 30s, brunette, soft features, wedding ring. Same kitchen forever. Same window. Same little blue mug on the counter every single pin. The mug is the moat. Every pin uses Google Flow to generate the same kitchen from a slightly different angle
Step 3: every pin links to an Amazon affiliate product. Cast iron pans. Wooden cutting boards. Pantry bins. Ceramic mugs. Spice racks. Each pin has a one-line caption written by Claude in the voice of "tired mom giving honest advice"
Step 4: he runs 14 of these "mom" accounts. Different aesthetic (clean girl, farmhouse, dark academia, Mediterranean, midwest, coastal). Same character file underneath. Different mug, different lamp, different cutting board. Each one is a separate Amazon Associate account routed through 14 different LLCs to avoid Amazon's "single operator" detection
Step 5: every pin he posts gets between 4,000 and 80,000 saves in 30 days. Pinterest pushes high-save pins to 1M+ users organically. 3% click rate. 6% Amazon conversion. 4-8% commission. The math compounds every single day forever because Pinterest pins don't decay like Instagram or TikTok posts. A pin he posted in March 2024 is still earning him $1,400/mo today
Stack:
$99/mo Google Flow
$20/mo Claude
$14 Wyoming LLC x 14 = $196 one-time
14 Amazon Associate accounts, all clean, all routed through different IPs + bank accounts
Pinterest scheduler $39/mo
Total cost: $158/mo
Revenue: $312,000/mo
Margin: 99.9%
He has never posted his face on the internet
He never will
His parents think he runs "a logistics company"
The Pinterest affiliate moat closes in 18 months when Amazon figures out the multi-account routing & forces selfie verification
Until then it's the most boring free money loop on the internet & nobody is allowed to talk about it
You're paying Meta $48 per install for a sleep app
Stop
Don't be fucking stupid
There's a Reddit playbook that's pulling installs at $0.83 each and 95% of consumer app founders have never heard of it because the agencies they hire would lose their entire retainer if they admitted it works
Here's the entire system:
Reddit has 138,000 niche subreddits with active daily users who self-identify as a buyer cohort. r/insomnia. r/perimenopause. r/dadjokes. r/sciatica. r/ChronicPain. r/RVliving. Each one is a pre-warmed audience that pays $0 to access
The play is stupid simple:
Step 1: Claude scrapes the top 200 posts of all time from any target subreddit. Identifies the 14 recurring pain patterns the community shares. Outputs a list of every emotional hook the audience has self-described in their own words
Step 2: I write 50 "first person story" posts using those exact patterns. Each post is 280-440 words. Sounds like a real user. Ends with one line: "I tried [my app] last month and it actually worked. Not sponsored, just sharing." Soft CTA, nothing salesy
Step 3: I run the posts through 30 aged Reddit accounts I bought on a dark forum for $40 each. Each account has 2-4 years of organic karma in unrelated subreddits. Mods cannot tell the difference. The posts hit the front page of the niche subreddit. Each one stays up for 11-72 hours before mods catch on. By then it's done its job
Step 4: every post drives 800-4,200 clicks to a landing page. 14% install rate. $0.83 cost per install when you blend post creation cost + account cost + occasional ban replacement
Step 5: I repeat in 14 different subreddits per app. Stagger across 30 days so no single sub sees more than 2 posts per week. Looks organic. Mods rarely connect the pattern
Last 90 days, one of my apps:
- 211 Reddit posts seeded across 14 subreddits
- 1,164,000 clicks
- 163,000 installs
- $135,000 total cost
- $0.82 CPI vs $48 CPI on Meta
I have a friend doing the same playbook for a perimenopause app. He's at $440,000 MRR. He has never run a Meta ad in his life
The "growth team" at your favorite app spends $180,000/mo on Meta and still can't beat one kid on Reddit with 30 aged accounts and Claude
Don't tell anyone
(don't actually)
I pray Apple keeps approving consumer apps
Every app I push live is worth $30k-$200k/mo and they have no idea who I am
They think I'm a "small US developer"
I'm 6 fake LLCs, 4 cloud phones, and a single Wyoming mailbox
Here's the part nobody on app Twitter is allowed to say out loud:
The App Store is the cheapest unregulated marketplace on earth right now. Apple gatekeeps the "creator economy" so hard nobody noticed they left the entire $180B consumer app industry wide open. Submit an app. Get approved in 36 hours. Run AI UGC at it. Print money. Repeat
Most "App Store success stories" you read about are 22 year olds running 6-12 apps simultaneously under different LLCs. They never appear on a podcast. They never go on stage. They never post their faces. They invoice through Stripe Atlas + a Wyoming registered agent + a virtual mailbox in Cheyenne
The reason they stay anonymous is simple:
Apple's 30% cut is the highest tax on earth & once your app crosses $1M ARR they will start randomly rejecting updates, flagging you for "deceptive marketing," and asking for "additional verification." Every operator at scale has a backup LLC in the queue, a duplicate codebase ready to relaunch under a new bundle ID, and a separate Stripe account warmed for 90 days
The infrastructure:
App 1: live, $42k MRR, primary LLC
App 2: live, $18k MRR, secondary LLC, different bank
App 3: in review, fresh LLC, new developer account, new bundle ID
App 4: in development, ready to submit if any of the above get yanked
The trick most people miss: every app is the same app. Same code. Different paywall language. Different icon. Different "for" audience. Sleep app, sleep app for moms, sleep app for shift workers, sleep app for veterans. Same backend, 4 listings, 4x the App Store search surface
Apple will eventually catch you & yank one. You laugh, swap the bundle ID, relaunch the next morning. The downtime is 14 hours. Revenue impact: $0
I have 3 friends running this exact structure. Combined MRR across them is $640,000/mo. None of their parents know what they do for a living. None of them have ever been on a podcast. None of them ever will
Apple cannot stop this & neither can you
I pray brands keep hiring real creators
Every $14,800 invoice they sign is a $14,718 profit margin with my name on it
They don't know one kid + Claude + 4 cloud phones can spit out 180 hooks before the "creative director" finishes his oat latte
The agency keeps $9,800
The "creative director" keeps $2,800
The editor keeps $400
The strategist keeps $600
The girl saying "wait i'm OBSESSED" on her unmade bed keeps $1,200
= $14,800 for one tiktok. Two revisions. She ghosts you in week 3 for "personal stuff"
My version of the same offer:
$99/mo Google Flow (face + scene)
$99/mo ElevenLabs (cloned a 58yo lady I found on Fiverr for $400, signed release, mine forever)
$20/mo Claude
$180/mo cloud phones + US residential proxies
$14 Wyoming LLC, one time
Cost per video, fully loaded: $0.62
Variations a month: 180
Revisions: 0
Usage rights drama: 0
Risk of her posting for your competitor next thursday: 0
I'll probably regret posting this but the brand markup on UGC is the dumbest arbitrage on the internet right now
A founder pays $14,800
The actual girl gets $1,200
The girl spends 18 minutes of her life
The agency keeps the rest for "creative strategy"
Meanwhile I'm in a Tim Hortons running 14 fake people on my macbook, billing through a wyoming llc, eating a $4.20 BELT
The brand opens their shopify at the end of the month and the $0.62 video outperformed the $14,800 one 41x
They don't fire the agency
They keep the invoice on the books because the board wants to see "brand-safe creator partnerships" in the deck
So I keep eating shawarma and stacking cash
Don't hate the player, hate the file
You're about to pay an agency $14,800 for one TikTok
Stop
Don't be fucking stupid
You can spin up 14 fake creators tonight for less than the agency charges for their "kickoff call" and outperform their video 41x by Sunday
Here's the exact stack the operators clearing $30k+/mo are running while you're on Zoom with a "creative strategist" who has never opened the TikTok app:
ACCOUNT INFRASTRUCTURE (the part nobody sells you)
Every account on its own cloud phone instance. Android. Dedicated US residential proxy locked to one state. Fresh device fingerprint. US Google account. US Play Store region. ~$4.20/account/month
The test most operators skip and burn $4,800 on:
Open the TikTok signup screen on the phone. Does it show +1 country code by default? No? Your geo chain is leaking. The account is dead before it posts. Nobody tells you this. I've watched grown men burn six figures on Seedance credits before catching it
WARMUP (168 hours decides everything)
Day 1-3: watch only. Scroll the niche 10-15 min, 3x/day. Like a few. Comment occasionally
Day 4-7: 1 post/day at the account's local peak window
Day 8+: scale to 2-3/day
The algorithm decides what KIND of account you are in the first 168 hours. Miss that window and you're cooked permanently. You cannot rewrite the categorization later. The pure-AI tool sellers will never mention this because their entire business depends on you skipping it
POSTING CADENCE
Stagger uploads 30-90 min apart across the whole farm. 40 accounts uploading at the same minute reads as a coordinated network. Distribution dies instantly. Looks like 40 strangers all posting good shit when you stagger correctly
THE COMMENT FUNNEL
Every video ends with a specific keyword:
"comment APP if you want the setup I run on this"
Keyword triggers the DM automation. The comment itself feeds the engagement signal the algorithm weights. Two birds, one CTA. Without this funnel your 2.5M views convert at 0.2%. With it, every qualified lead drops into a 3-message DM sequence into the offer
THE 73/27 RATIO
73% AI, 27% real human seed footage. Pure-AI dies inside 14 days. Every time. TikTok runs a 4-second synthetic detection pass on every upload and pure-AI fails it. Same face geometry across 50 accounts. Same micro-motion signature. Same upload cadence per IP block. The pattern is too clean and the algorithm strangles you in the cradle
The fix: I pay a girl in the Philippines $5/video. iPhone. No script. 3-4 seconds of lifestyle b-roll. She makes 200 clips a week. The seed footage gets spliced into the cold open of every AI video. The algorithm reads it as the account's biometric signature. Every subsequent AI clip inherits it. Account stops getting bucketed as synthetic
THE MATH
280 videos in 14 days
3,182,000 views
18,400 installs
$32,400 MRR added
Total creative spend: $172
The agency burned $14,800 to make one video that got 8,200 views and a "let's iterate" email
The founders winning this aren't smarter than you
They just stopped being polite
Now you know
It's funny how you could quite easily make $200k+/mo selling sleep tea, collagen, and ceramic mugs to middle aged women using AI characters they will never meet, never miss, and never realize do not exist
The grandma niche alone is a $4.8B retail category and 90% of operators are still trying to find a "real" grandma to put on camera
Don't be a retard
You can build one in an afternoon
Pinterest is full of broke aesthetic women who would kill for $1,800/mo to be the face of an offer. Fiverr is full of 60 year old women with rural accents who will sell you a voice clone for $400 one time. The combination of those two assets is more valuable than every "creator partnership" the average DTC brand has ever signed
Here's the playbook everyone's sleeping on:
Step 1: scout one Pinterest girl with the exact aesthetic. Clean kitchen, soft tones, looks like someone's older sister. DM her $1,800/mo to be a static reference (no posting, no video, no risk to her)
Step 2: train Google Flow on her face + room. Lock the room from frame one. Same kitchen, same lamp, same towel in the background every reel. The towel is the moat. People in comments will start asking about the towel within 30 days. That's how you know it's working
Step 3: voice. Find a real 58yo woman with a slight rural accent on Fiverr. Pay $400 for 60 min of varied audio (12 min natural conversation, 12 min product demos, 12 min CTAs, 12 min emotion variations, 12 min single-word delivery). Clone it. Use forever
Step 4: caption style. Written like she typed it in a Walmart parking lot. Lowercase. Missing apostrophes. One typo per caption on purpose. Polished captions kill the frame instantly. This is where 95% of AI operators get blown up
Step 5: post the same 5-reel emotional arc:
- "my daughter said nobody would buy these..."
- workshop b-roll (also AI)
- first order packed
- "i made $1,200 this week and cried"
- "thank u for changing my life"
Step 6: alibaba > shopify. $3.40 unit cost. $34 sale price. 68% margin after meta + shipping. Subscribe & Save annuity adds 30% LTV
A reel posted at 9:08pm to women 45-58 in midwest states converts 6x better than morning posts. That cohort is the highest LTV buyer in retail. 50M women in the US. Repeat purchase rate 71%. Refund rate 2.1%
You don't need a face
You don't need a real grandma
You don't need a "brand partnership"
You don't need a marketing degree
You need a file, a voice, a room, and a 5-reel arc
I have friends running 14 of these characters off one cloud-phone farm and a single shopify backend. One of them cleared $2.3M last year selling ceramic mugs. He has never met a single one of his "creators." None of them have ever existed
The market doesn't care if the person is real
The market cares if the story is specific enough to feel real
Run the numbers
The most expensive purchase a 28 year old can make in 2026 is a real job
The white collar prestige economy is structurally a delayed-trap pricing model and almost nobody is doing the actual math
A $340k/yr goldman job sounds like the apex of the meritocracy. After taxes, NYC rent, the lifestyle cost of "being a goldman person," and the time cost of 80 hour weeks, the effective hourly rate is roughly $58/hr
A $980/hr small business owner in atlanta is generating his income through ownership, not labor. His $1.8M in distributions are taxed at capital gains rates, not ordinary income. His lifestyle cost is 1/4 of NYC. He's running 35 hour weeks
After-tax, after-lifestyle, after-time, the small business owner is 14-20x richer per hour
The math has been true for 6 years. The structural reason it's getting worse is that AI is compressing the value of labor and amplifying the value of ownership
A goldman analyst's labor is becoming cheaper to replace every quarter. His value-add is bounded by what a model can't yet do, and that surface area is shrinking at 30-40% per year. His salary will not rise at the rate of inflation in the next 5 years because the marginal labor he provides is competing with AI agents that get cheaper every month
The small business owner's ownership is becoming MORE valuable every quarter. His business is generating more profit on the same revenue because AI is compressing his operational costs. His exit multiple is going up because PE firms are paying premiums for AI-enabled SMBs. His passive income is rising
The status trap is the part nobody talks about
The goldman analyst job has 4 specific addictive properties that lock you in even when the math goes negative
1. The prestige compounds INTERNALLY (you become a VP, then MD, then partner) but doesn't translate externally. Leaving means starting over in a category where the prestige doesn't exist
2. The lifestyle costs lock in. Once you live in tribeca, drink at $200 wine bars, and vacation in capri, going back to a midwestern lifestyle feels like a status downgrade even if it's a financial upgrade
3. The peer network locks in. Your friends are all goldman/blackstone/citadel. If you leave, you're not just leaving a job, you're leaving the social structure your 20s were built around
4. The cognitive frame locks in. Goldman teaches you to think in deal cycles, quarterly compensation, and bonus pool calculations. The frame of an entrepreneur (multi-year compounding, customer-led iteration, asymmetric upside) is structurally incompatible. The longer you stay the harder it is to retrain your brain
The trap is that all 4 lock-ins compound with tenure. A 24 year old goldman analyst can leave with no cost. A 28 year old VP has 2 of the 4 locked in. A 33 year old MD has all 4 locked in and will essentially never leave
By the time you cash out, you've spent the years of compounding capability working for someone else's compounding capital
The 24 year olds entering goldman this september are buying a $58/hr job that destroys their 20s in exchange for the right to put 4 letters on their linkedin
For 50% of them it's worth it. For the other 50% it's the most expensive purchase they will ever make and they don't know it yet
few
Your dad's small business is suddenly making 4x more money this year because he learned how to use AI before you did
The structural advantage of legacy small businesses in the AI era is one nobody is pricing in
A 54 year old HVAC operator has 11-30 years of operational records sitting in filing cabinets, quickbooks files, and email threads. Every job. Every customer complaint. Every part replacement. Every diagnostic decision. Every successful close. Every failed bid
This is the training data nobody else can buy
The 24 year old SaaS founder building "AI for trades" cannot acquire this data. He doesn't have it. His startup costs $4-8M to seed it would take to scrape industry-wide data, and the scraped data is noisy because it pulls from every operator at every quality level. He ends up with a generic model that performs at the 50th percentile of the industry
The 54 year old HVAC owner has DECADES of proprietary data on his specific local market, his specific customer base, his specific competition, and his specific failure modes. When he feeds this into claude and asks "find patterns i missed," he gets insight a generic model literally cannot produce
His close rates jump. His routing efficiency jumps. His marketing copy outperforms the local agencies because his copy is calibrated to the actual symptoms his customers describe instead of the industry-standard language the agencies recycle
The agencies cannot compete. The SaaS startups cannot compete. The 24 year old MBA graduates cannot compete. The advantage is in the data and the data only exists in the legacy operator's archives
The high IQ play the operators are running is small business ROLLUP
Take your AI-enabled operation that does 4x what a normal HVAC company does. Use SBA loans to acquire 2-3 adjacent companies in nearby counties. Install your AI dispatch and lead scoring at the acquired companies. Their economics jump 3-5x within 6 months. You bought them at 3-4x EBITDA. They're now worth 8-12x
Repeat 8-15 times over 3-5 years. The aggregated entity is now a regional HVAC platform doing $40-80M in revenue. PE firms specialize in acquiring exactly this profile because home services is the most fragmented and least technologically advanced industry in the US economy
The exit multiples on aggregated home services platforms run 12-18x EBITDA
The 54 year old doesn't tell anyone what he's doing because the AI angle is the unfair advantage. If the other operators in his county figure out the same playbook, the acquisition prices go up and his rollup math collapses. He maintains the appearance of being a normal HVAC owner buying small companies for "consolidation reasons"
In 2027-2028, business press will write articles about "the surprising consolidation of home services" and miss what actually happened. They'll attribute it to "private equity interest" and "demographic shifts in skilled trades labor." They'll completely miss the fact that the consolidation was enabled by 50-60 year old operators who quietly adopted AI before the analyst class noticed
The most valuable AI applications in the next 5 years are not in san francisco
They are in 6,000 small businesses across the midwest run by operators in their 50s who have never posted on twitter
he he he haw
3 quarters away from the first publicly disclosed case of a company being fully run by 1 person pretending to be 40 people. The worst part is the company is probably going to be VC funded by the time we find out
VC diligence is structurally optimized to verify revenue, not to verify humans
The standard $4M seed diligence process checks: cap table, customer references, founder backgrounds, financial controls, IP ownership, market size. All of these can be faked by 1 operator running a full identity stack
Cap table: split equity across 6 delaware LLCs you control under aliases. The cap table shows 6 founders. You own all 6 entities
Customer references: pay yourself through 4 different LLCs to be a "customer." Each LLC has its own bank account, its own EIN, its own website. The VC interviews "customers" who are technically separate legal entities. The wires they make to your company are real. The legal liability they assume is real. The revenue is real
Founder backgrounds: midjourney for headshots. ChatGPT for bios. Linkedin profiles warmed for 4-6 months before pitching. Podcast appearances faked by submitting "guest pitches" to small shows and reading from a script under aliases
Financial controls: real bookkeeping. Real CPA. Real audit trail. The money is real. The accounting just routes between LLCs you control
IP ownership: your fake co-founders sign assignment agreements. Each fake co-founder is a real legal entity. The IP transfers are real
The entire structure is legally bulletproof because every individual component is legitimate. The fraud isn't in any single piece. The fraud is in the COMPOSITE. And the composite isn't illegal anywhere in the US because there's no statute that says "you can't be 6 different LLCs at once"
The IRS doesn't have a framework. The SEC doesn't have a framework. State business filings don't have a framework. The whole concept of "person identity" in corporate law assumes 1 human = 1 person, and that assumption broke 14 months ago
When the exit happens (typical 4-7 year timeline), the operator dissolves the fake employee LLCs in sequence. Each dissolution triggers a separate capital gains event in delaware (no state tax). The total tax burden on a $40M exit drops from roughly 37% federal + 13% state down to 23% effective blended rate. The savings on the exit alone run into the 7 figures
VCs will eventually figure this out. When they do, they're going to demand "human verification" as part of diligence. Bio scans, in-person meetings, embedded operators on the cap table
The catch is by then, deepfake video will be indistinguishable from real video in real time, voice clones will pass turing tests in any language, and the "human verification" will be done by clones verifying clones
We are 18-30 months from the first $100M+ company being acquired with no humans on either side of the transaction
navigate
If AI just replaced the actor IN the ad, the next thing it replaces is the buyer OF the ad
T(his sounds like a tweet from a guy who reads too much techcrunch)
It's happening at like 6 fortune 500 companies right now and the procurement teams running the experiments automated their own jobs and didn't tell anyone
I had drinks with a VP of procurement at a top 40 fortune company in march. Won't say which one...
He showed me an agent he built over a weekend
It identifies vendors. Runs them through structured discovery. Scores them. Writes the procurement memo. Joins the sales calls. Asks the questions. Rejects the unfit ones. Forwards the 2 finalists to him
He said he hasn't done his actual job in like 9 months
His team of 13 people don't know. His director doesn't know. His CFO definitely doesn't know
The only person who knows is his wife and she thinks it's "kind of dishonest" but they just put a down payment on a place in tribeca so she doesn't bring it up
I asked him what happens when his director finds out
He said "they won't. They don't have the technical literacy to find out. They'll just notice procurement is closing faster and assume i'm a good manager"
The high IQ part is what he doesn't tell anyone
He doesn't post the agent on github. He doesn't put it on his resume. He doesn't mention it in his 1:1s. He runs it through a personal AWS account billed to a delaware LLC he set up under his wife's maiden name. The cloud costs don't show up on the company's books. There's no paper trail at his employer
If they ever did find out, his cover story is rehearsed. He'd say it's a "personal productivity tool i use for organizing my own notes" and the org's compliance team doesn't have the sophistication to disprove it
He's not stupid. He's running an asymmetric play with hidden infrastructure and zero legal exposure
I asked him what he does all day. He said he goes to barry's at 9am and reads. Answers maybe 4 emails. Sits in a conference room and "thinks." Has dinner with his wife
He's the highest paid person in his peer group and the only one doing like 12 hours of real work a month
I don't know what to do with this information either
I think there's like 80,000-100,000 of him right now in mid to senior procurement, ops, and middle management roles across north america. Quietly running agents. Quietly collecting salaries. Quietly knowing the org is going to collapse in 18 to 24 months
They're waiting for the music to stop and trying to save as much as they can before it does
The entire white collar economy is now a slow motion BANK RUN and the depositors are pretending they don't see it
navigate
The most underpriced asset in the next 4 years is being INVISIBLE
I've watched 3 friends get publicly destroyed in the last 13 months for things they posted under their real names 4 to 8 years ago
One lost a $3.8M deal
One got removed from a board seat
One is in the middle of a lawsuit I can't talk about (still)
The rule everyone was taught from 2010 to 2024 was "build a personal brand. Attach your face to your work. Become the trusted voice in your category"
That rule died
It died because the cost of being attached to a name in a public moral panic is now infinite and the cost of being anonymous is approaching zero. You can run a dozen LLCs out of delaware. Sign contracts under aliases. Take 6-figure wires into bank accounts that trace to nothing
The smartest operators I know are quietly erasing themselves
One of them deleted a twitter account with 91k followers in february
One pulled 13 podcast appearances offline
One filed a trademark withdrawal on his own fucking name
These aren't paranoid niggas. These are some of the most successful operators in their categories. They all came to the same conclusion at roughly the same time, independently, in the last 18 months
Here's the part that ai can't tell you
It's not just about lawsuit exposure or moral panic. The deeper play is that anonymity restores the ability to switch faces
A personal brand locks you into one identity. You become a 1-product business and the product is you. When the market moves you can't pivot because your audience expects you to keep doing the thing you're known for
An anonymous operator can be a credit repair guy on monday, a real estate fund on tuesday, a wellness brand on wednesday, and a b2b consultancy on thursday. Same brain, different masks, different LLCs, different bank accounts, different tax strategies for each one
The personal brand operator is locked into one tax structure, one identity, one professional reputation, one set of regulatory exposures. The anonymous operator is running 9 parallel businesses with no narrative obligation between them
When one mask gets retired (regulator, lawsuit, scandal, market shift), the other 8 keep running. The operator just deletes the mask and walks away. No press cycle. No reputation damage. Nothing to defend
This is the real game and almost nobody under 30 understands it because the entire "build in public" generation was trained to believe transparency is leverage
Transparency was leverage for 14 years because the platforms rewarded it and the regulators were slow
Both of those conditions reversed
I'm writing this from an account that doesn't have my real name on it, and that's intentional. 4 years from now you won't be able to find me at all, and that's also intentional. If you're smart you're 12 months away from making the same choice and you don't know it yet
The next 200 multi-million dollar fortunes in the consumer category are going to be made by mfs you've never heard of and NEVER WILL, and they're going to laugh about it on a beach in cape verde while the personal brand crowd is on podcast number 91 trying to "stay relevant"
The highest status play in 2030 is having no public footprint at all
start now
Had a call last week with an old friend who runs growth at a $174M consumer brand. Won't name him
He asked me how my consumer app was doing
I told him roughly $34k/mo on a content budget of $2,900
He didn't say anything for about 6 seconds
Then he said "we're spending $510,000 a month on creative right now"
I asked him what that buys him
He said 41 ad variants
I did the math out loud on the call
His per-video cost: about $12,400
Mine: like $9
The exact multiple doesn't fucking matter, the point is the gap is so wide you can't even talk about it in polite company
He stared at the camera like he just realized his pants were on backwards
I tried to soften it. Told him brand standards are genuinely different at his scale. Told him there's a real argument for human-led creative when AOV is $180+. Told him his agency is reputable. Told him board-level brand decisions can't be delegated to a 22yo in a basement
I was lying about most of it. I was trying to get him off the call before he had a fucking breakdown on camera
He asked me if I'd talk to his board
I told him I can't. Because his board is going to interpret what I tell them as a reason to FIRE HIM
We got off the call
I don't think he's going to make it
Here's the part nobody is saying
His agency contract has a 14 month minimum and a 6 month notice period. So even if he walked into the CEO's office tomorrow and said "we're getting robbed," the brand is contractually locked into the next $9M of agency spend whether they like it or not
His agency knows this. His agency's contract was written by a lawyer who specializes in "marketing services agreements" for a reason. The 14/6 structure is the agency's actual moat. The creative work is the loss leader. The minimum term is the product
Every agency-locked brand on earth is in the same trap right now. The math is brutal. The exit is contractually delayed. The boards are unaware. The CMOs are paralyzed. The agency owners are riding it out until the next renewal cycle and then they're going to retire to greece
I'm going to keep my mouth shut and keep running my app
If you know a CMO at a $50M+ brand, hug them
most are not gonna make it
The funniest thing about running a "company" in 2026 is you can fully fake having a staff and nobody at any level of corporate procurement will catch you
I have a friend running a b2b agency out of a one bedroom apartment in lisbon. Won't name him because he'd actually kill me
His company has 19 employees on the website
A CEO named marcus. A head of strategy named lena. A head of operations named raj. A senior designer named priya. A junior copywriter named tom. Etc
None of these niggas exist
He generated all 19 headshots in midjourney 9 months ago, wrote the bios himself in an afternoon, registered the linkedin profiles through residential proxies and warmed them slowly over 5 months. They all post once a week. They all share each other's articles. They all attend "team retreats" in cape town that consist of him uploading 1 stock photo and writing 4 captions
His clients have never met any of them. His clients pay him $74,000/mo
Last month one of his biggest clients asked if "marcus the CEO" could join a strategy call. He said marcus was on PTO. Client said no problem can lena join instead. He said sure
3 hours later he hopped on a zoom with a deepfaked avatar and a voice clone he set up the night before
The client LOVED lena. Asked if she could lead the next phase of the engagement
He's been "lena" for 8 calls now. He charges 27% more when she's on the project because the client thinks "the senior strategist is involved"
The high IQ part is that he picked the names with actual intention
Marcus is an anglo name because the buyer is more likely to assume competence
Lena is european enough to imply formal training but not foreign enough to trigger an accent objection on a zoom call
Raj is south asian because the buyer expects the head of operations to be south asian (this is the unspoken bias in every procurement diligence sheet and he reverse engineered it)
Priya the designer is the same logic, the buyer doesn't question a south asian senior designer
Tom the junior copywriter is the cheapest name in english, conveys "young and disposable" which lets him justify the labor cost line item
The roster isn't 19 random fake people. It's 19 fake people picked specifically to defeat the unconscious bias of the procurement officer reviewing the vendor sheet
He told me the vendor diligence process at fortune 500 companies hasn't been meaningfully updated since 2014 and he intends to ride it until 2031
I don't know what to do with this information
I think we just opened pandora's box and nobody is going to close it
Nobody is willing to admit what just happened to marketing in the last 11 months
A year ago the cheapest creative budget at a real consumer brand was $80,000/mo for 14 videos. Today the actual cost to ship 280 tested variants is $214 in software and 4 hours of attention from a 22yo with a refurbished samsung
That's not "marketing got more efficient." That's an entire fucking industry getting NUKED while the executives sit in q3 planning meetings discussing brand voice frameworks
Every CMO making $314,000 right now is sitting in a chair priced on a cost model that died 11 months ago and not a single one of these mfs is going to bring it up because they have a kid in collegiate school and a mortgage in fairfield
I know one of them. Won't say his name. Runs marketing at a $174M ecom brand. Last week he asked me, half joking half not, "do you think i have 18 months"
I told him probably 13
He didn't laugh
The brands that survive the next 2 years are not the ones with the best strategy. They're the ones whose CMO walks into the CEO's office tomorrow and says "fire 9 of my 14 people and let me run the department like a kid in a basement"
Zero of these executives are doing that
The boards aren't going to force it either because the board members got their seats by being the kind of person who hires a $314,000 CMO in the first place. The whole approval chain is the problem and the people in the chain are the ones who would have to dismantle it
I'm watching this play out in slow motion across about 16 brands I'm close to. All of them know. All of them are paralyzed. The holding companies that own them are about to make a generation of acquisitions at like 0.3-0.4x revenue and the consultants pitching the deals will pretend they saw it coming
The next 2 years of mid-market consumer brand acquisitions are going to be the most violent wealth transfer in the category since the 2009 distressed cycle. The acquirers will be holding companies run by mfs in their late 20s who never had a marketing department. The targets will be CPG brands their parents bought from for 30 years
And nobody in business press is going to write about it accurately until 4 years after it ends, because the journalists covering consumer brands are friends with the CMOs about to get sold for parts
I'll probably regret posting this
few
You could probably make $124,500/mo selling boring contractors a private feed of government buyers before the buyers even know who to call...
Procurement portals are ugly as fuck
https://t.co/sZcdRpFmnj, school district board minutes, local council PDFs, city budget docs, all these horrible little websites that look like they were built by a divorced IT guy in 2006
But inside them is buyer intent with money already approved
A school district posts that it needs new visitor check-in software
A city posts that it wants fleet cameras
A public library posts that it needs cyber training for staff because Barbara clicked one phishing email and now everyone is panicking
Every local vendor in that category should know within 10 minutes
They don't
They're busy driving around in a van eating gas station sandwiches wondering why referrals are slow
The play is stupidly obvious...
Pull public docs daily
Turn government goblin language into plain English
Send vendors a private alert like:
"Plano ISD is looking for visitor management software. Budget range looks like $80k-$120k. Decision meeting on Tuesday. Contact page here. Previous vendor attached."
Charge $249/mo for the feed
500 vendors = $124,500/mo
Then upsell proposal writing, intro brokering, competitor tracking, "we found 18 contracts you missed last month" reports, all that boring stuff that makes grown men pull out company cards
Government websites look like diabetes but the budgets are real
Someone is going to wrap public procurement data in a nice dashboard and become the Bloomberg Terminal for roofers, HVAC guys, security installers and random B2B vendors
Not sexy at all
Which is why it will print
Amazon sellers are throwing away the customer after the first sale because they're too lazy to bribe buyers with a stupid warranty QR code...
Amazon hands you the sale
You use the QR to get the relationship
That's the whole game
Imagine you sell a dog nail grinder for $16.99
Most sellers just ship the product, beg for a review, and then cry because Bezos owns the customer
Smarter play:
Put a little card in the box
"Register your 2-year warranty + get a free dog grooming checklist"
The buyer scans
You ask:
- Dog name
- Breed
- Age
- Main issue
- Email + phone
Now you don't have "a customer"
You have Molly, 34, Ohio, owns a 9 month old golden retriever called Biscuit who hates nail trims and chews the skirting board when left alone
That data is worth more than the nail grinder
Golden retriever puppy = training course, chew toys, calming treats, pet insurance affiliate, fresh food box, grooming subscription
Senior bulldog = joint powder, breathing support, vet discount club, soft bed, dental chews
Same product, different dog, different money path
You could build an entire backend from one boring Amazon front-end
$16.99 grinder gets the buyer
$29/mo grooming club keeps them
$89 video consult fixes the scary problem
$11.50 affiliate payout from the food brand stacks in the background
The funniest part is customers actually like doing it because "warranty" feels responsible
They think they're protecting their purchase
You're building a customer file that tells you how to sell to them for the next 8 years
If you're selling on Amazon and you're letting Bezos keep all the relationship data, you're getting cucked by a bald man in a cowboy hat
Fix up
Every viral "my landlord left black mold in my kid's bedroom" video is a lead farm wearing sad music...
The comments are insane
"Same thing happened to me in Phoenix"
"My baby had asthma for 2 years because of this"
"Can I sue if my landlord ignored emails?"
"Wait this is illegal?"
That's not a comment section
That's a room full of pre-qualified people asking for a claim form
You can build a disgusting local lead-gen machine around this without being some cringe law firm doing billboards with a guy pointing at a car crash
Pick one pain category:
Black mold
Wage theft
Botched filler
Bad dental work
Moving company scams
Nursing home neglect
Then build a dead simple "claim checker" page around it
"Upload photos, pick your state, answer 6 questions, find out if you might have a case"
Run TikTok content reacting to viral complaint videos, make the angle emotional, pin the checker, push people into intake
Lawyers already pay stupid money for qualified leads because their own marketing is usually a man in a blue suit saying "injured? call now" like it's 1998
You give them:
- City
- State
- Photos
- Timeline
- Landlord name
- Email proof
- Severity notes
$65 per qualified lead × 600 leads = $39,000/mo
And that's light if the niche is nasty enough
The darker the pain, the more people search in silence before asking for help in public
TikTok comments are full of broke people accidentally raising their hands
Someone is going to turn viral complaint culture into intake funnels for lawyers and quietly stack ridiculous cash
Sad music + legal intent = money
YouTube timestamp comments are poor people telling you which part of a free video should have been a paid product...
Go under any 2 hour podcast with a niche expert and you'll see the same thing:
"23:18 changed my life"
"41:02 is the part everyone needs"
"Does anyone have a template for what he said at 58:10?"
"Can someone make notes for this?"
The audience is showing you the exact moment where free content became valuable enough to package
They're also showing you how lazy they are
A 2 hour immigration lawyer interview might have 9 minutes where he explains the visa mistake everyone makes
Turn that into a $29 checklist
A mechanic podcast might have one section where he explains how to inspect a used car without getting finessed
Turn that into a $19 printable sheet + video vault
A dating coach rant might have one tiny section where women in the comments keep asking for examples
Turn that into a $39 opener bank with voice note reviews
Do this across 100 videos in one niche and you have a map of what people actually care about, built from the moments they replayed and begged strangers to summarize
No genius required
Just watch where the crowd points
The workflow is easy:
Find long videos
Pull timestamp comments
Cluster the repeated pain
Package the tiny answer
Make it look clean enough that people trust it
A $29 guide only needs 1,800 sales to do $52,200
Add a private Telegram, weekly call, templates, AI helper, whatever makes the buyer feel less alone
People don't pay because the information is impossible to find
They pay because finding it makes them feel like a rat in a filing cabinet
YouTube already found the demand for you
The timestamps are the receipt
@shipwithjay yesssir. the only way to figure out what people want is to test angles/videos on scale and have a system to be able to understand why some do well and some dont. Organic is a numbers game end of the day
TikTok is a casino and most consumer apps are walking in with 300 chips/month wondering why they lose.
That is the biggest thing I learned building our app to ~$300K ARR.
Here’s how to look at it:
Every video is a chip on the table.
Every slideshow is a free play.
The only way to win is to have more chips than the player sitting next to you.
And that player is your competitor.
They’ve been running for 6 months.
They can afford to spend $15K/month on 200–300 UGC videos.
They have creators, editors, agencies, ads, and a full content team.
When we launched, that was the market.
6+ competitors.
Some spending close to $20K/month on UGC alone, and we still came out on top.
We were not going to outspend them.
So we didn’t.
We built synthetic creators that act like real creators.
Different faces.
Different angles.
Different hooks.
Different accounts.
Now we get 3M+ views/week with 40 synthetic creators.
Our cost is ~$3K/month.
Production. QA. Account ops. Distribution.
All included.
We literally run a phone farm in Canada to make this possible.
Most people are still trying to make “good videos.”
We’re flooding the algorithm with so many angles, hooks, and synthetic creators that the customer has no choice but to keep seeing us.
Then eventually, they click.
Then they download.
That’s the game.