Operator · not influencer

Claude built it.
GPT pitched it.
Zuck sold it.

Seven months ago this operation didn't exist. Jan: started learning Claude Code. Feb: shipped the SaaS. Mar: GPT drafted the ads — first $57.19 to Zuck on the 12th. Apr: $102,431.19, pixel-attributed. Headcount: one.

CURRENTLY DEPLOYING ~$1,086/DAY · PIXEL-VERIFIED BASE AUG 27, 2026

Read the $100K Month — $19 Watch live — $99/mo

JAN · first Claude Code session FEB · SaaS shipped MAR 12 · first $57.19 to Zuck APR 30 · $102,431.19 month

SWAP: assets/portrait.jpg

Islamabad → Meta's auction · est. Jan 2026 · headcount: one

$165,805

handed to Meta · Mar–Aug 2026 · 3 accounts

$102,431

attributed in April alone · 3.87x · best month

2,470

customers pixel-counted · six months, three accounts

6,951,123

impressions bought · 244,471 clicks

Pick your door

Which one are you?

01 · THE BUILDER

You ship. Nobody buys.

Claude Code made you a builder. This is the selling half — rules, receipts, live spend.

START WITH THE $19 FILE → 02 · THE BUYER

Thresholds, not vibes.

Kill at $100/0. Statics 2.03x vs video 0.97x. Watch a live account enforce it.

ENTER THE WAR ROOM → 03 · THE FOUNDER

Customers, not courses.

818 buyers in 30 days at $32.38 each. The rules cost less than one of them.

READ THE $100K MONTH → 04 · THE STARTER

No product yet.

Day 1 was $57.19 and a cold pixel. Read how the whole thing began — free.

READ THE STORY ↓

The ledger, chronological

Receipts before reach.

No résumé section, no certificates wall. Seven months of dated receipts instead — the whole operation, from first tutorial to the auction.

Jan 2026

The first session

Opened Claude Code in January. The plan: build the product myself, with AI, in weeks. No dev team, no agency, no excuses.

Feb 2026

The product ships

A working B2B SaaS, built and launched in February. Vibe-coded end to end — the stack costs closer to $0 than you'd like.

Mar 2026

GPT writes the ads

First drafts by GPT, thresholds by me. March 12: $57.19 into a cold pixel. The machine doesn't know us yet.

Apr 2026

The $102,431 month

$26,489.59 in → $102,431.19 attributed → 3.87x → 818 customers. Month three of the whole operation. Documented day-by-day in The $100K Month.

May 2026 →

The doors open

The playbook becomes The $100K Month. The desk becomes The War Room. Build, buy, publish — now with an audience.

Now

$1,000+/day, in public

Build with AI. Buy from the auction. Publish the ledger weekly. You're standing in the middle of the experiment.

The current operation

One loop, running daily.

01 · BUILD

Ship in weeks

Built with Claude Code — AI development that compresses quarters into sprints. January: learning. February: shipped.

stack cost: closer to $0 than you'd like

02 · BUY

Rent the auction

$1,086/day average into Meta across three ad accounts. Tested at $50/day per ad set, governed by fixed kill-and-scale thresholds.

6,951,123 impressions · 244,471 clicks so far

03 · PUBLISH

Show the ledger

Weekly plans, verdicts, and results — cents intact. Highlights here; the full feed lives in the War Room.

plans post before results exist — on record

$57.19

Mar 12 · first dollar · cold pixel, zero history

11.6x

Apr 9 · $567.47 → $6,577.89 · 47 customers

$102,431

Apr 30 · month closes · 3.87x · 818 customers

$1,086/d

current pace · the loop hasn't stopped

Operating doctrine

Five rules I don't renegotiate.

Thresholds, not tantrums

Ads die at fixed dollar lines — $60, $100, $150 — and graduate at proof, not hope. The full law lives in the products; the discipline is free to copy.

ii.

Receipts beat reviews

No testimonial wall on this site. Pixel-attributed exports, dated and republished weekly, out-argue any five-star quote I could collect.

iii.

The boring creative wins

On my accounts, receipt-style statics run 2.03x against video's 0.97x. I stopped arguing with the data; my ads look like invoices now.

iv.

Don't touch the machine

~50 budget edits in 60 days once reset my learning phase into oatmeal. Schedules move budgets; moods don't.

Publish the ledger

Numbers to the cent, dated, before anyone asks. Credibility is arithmetic, not adjectives.

The work, packaged

Two artifacts. No upsell maze.

The record · $19 once

The $100K Month

April 2026, reconstructed decision by decision: $26,489.59 → $102,431.19, pixel-attributed. The testing grid, the six thresholds, the copy codex, the creative doctrine — and the fade protocol that guards winners.

The 30-day ledger, annotated

All six kill / scale rules with their scars

The vault — 100+ campaign lessons from six months of daily fire

$19 one-time

Read the month — $19

INSTANT ACCESS · 30-DAY REFUND

The live feed · $99/month

The War Room

The same operation, watched live: Monday plans before results exist, Thursday kill/scale verdicts with thresholds on screen, Sunday ledgers with losses intact — on a members platform that's itself a vibe-coded SaaS.

3 drops weekly from $30K+/month of live spend

Community, Q&A, and the full archive

Predictions on record — grade me against my own calls

$99 /month

Enter the War Room — $99/mo

CANCEL ANYTIME · ONE CAUGHT AD SET COVERS SIX WEEKS

The console

Ask the ledger.

operator@ranasohaib — the ledger speaks in numbers

LEDGER OS v3 · pixel-attributed · type help and press enter.

EVERY ANSWER IS A REAL EXPORT. THE CONSOLE DOESN'T DO ADJECTIVES.

One operation · four doors

The same ledger wears three masks.

This site is the operator's desk. The spectacle, the letters, and the courtroom each get a door of their own — same numbers behind every one.

WEPAYZUCK.COM

The Tribute Ledger

"I pay Zuck $1,000 a day. Watch what he does with it." The counter, the receipts, the spectacle.

VISIT THE LEDGER →

DEARZUCK.COM

The Letters

Every Sunday, a letter to the man who never replies. Gave / Got / Learned / Next — cents intact.

READ THIS WEEK'S →

KILLORSCALE.COM

The Courtroom

Every campaign gets one verdict. 40 trials this year: 6 scaled, the rest died on schedule.

ENTER THE COURT →

Asked often

Plain answers.

Are the numbers audited?

They're Meta-pixel-attributed conversion values from Ads Manager, exported via Windsor.ai on Aug 27, 2026, from accounts I operate. Attribution isn't bank profit — it's the standard working metric of paid social, and I publish it with that label every time.

Which SaaS is it?

Deliberately unnamed — keeping the name out keeps the numbers the point (and the ads unbotherable). Members see enough context to judge everything that matters.

Why sell a $19 file if the ads print?

Because documentation compounds. The spend is the story; the story pays its own postage. And no — you don't need me for clients. The file carries the doctrine; the Room shows it running.

Will your system make me a $100K month?

No promises — run from anyone who makes them. You're buying my documentation and my rules, which make experiments cheaper and verdicts faster. Your offer, product, and discipline decide the rest.