Текст заказа
This isn’t another growth job
Most growth teams call themselves data-driven. They have a dashboard, an agency and a ChatGPT licence to write ad copy faster. That’s the 99%.
We want to be part of the 0.1%: a company where AI doesn’t just help the growth team, it runs the growth loop. Every day, on every channel.
Then it’s your chance to define a new category : The Operating System (OS) for Human Greatness Augmented with AI.
Science-backed. Design-led. AI-powered. Global.
We’re not looking for someone who only advises on what to test, we’re looking for someone who builds it. A hands-on Head of Growth who can ideate, write the copy, design the strategy and launch experiments directly.
WHAT AI-NATIVE GROWTH MEANS AT PARADOX
Our growth machine has to be able to:
1. Generate hypotheses. Read the signals (ad performance, webinar attendance, sales calls, objections, churn reasons, comments, competitors’ ads) and propose what to test next, with a cost and an expected impact.
2. Act across every tool. Not a chatbot in a tab: it reads and writes in our whole stack, from ad accounts to CRM, landing pages, webinars, email, social DMs and analytics.
3. Hold the full context of Paradox. Our DNA, offers, voice, member stories, past campaigns and what they taught us, in one shared company brain.
4. Ship its hypotheses. Produce the hook, the ad cut, the page, the email sequence or the funnel variant, and put it live.
5. Measure with real KPIs. Every test is tied to a number that matters (cost per qualified lead, show-up rate, close rate, CAC, payback, LTV), not to likes or reach.
6. Learn and loop on its own. Scale what works, kill what doesn’t, write the lesson back into the brain, and launch the next cycle without waiting for a meeting.
This is already real. Our growth hub tracks every hypothesis and experiment, with its result. An ads library of thousands of creatives feeds our creative research. Agents cut ads from long-form content, prepare webinars, wire Instagram comment-to-DM automations and audit our tracking after every live. Athena, our AI coach, works with members every day.
OUR MISSION
At Paradox,
We believe self-help is broken: generic theory, a quick high, no real change.
We believe people deserve more. More than content. More than motivation.
They deserve a system that knows them, one that adapts, challenges, and grows with them.
Growth at Paradox means bringing people into something that actually changes their life, and proving it.
We do that by merging :
- Human Intelligence (coaches, peers, and mentors)
- Structured Intelligence (protocols, mental models, rituals)
- Technological Intelligence (AI personalization, feedback loops, adaptive journeys)
To create one seamless ecosystem: Paradox OS.
Short history
Paradox was founded by David Laroche, a creator followed by over 2 million people and known for his interviews with icons like Tony Parker, Mike Horn, Richard Branson and Gad Elmaleh. We started in French. We’re building for the world.
Here’s the bet: a company that learns by itself grows in a way a company of people alone can’t. A small growth team running a self-learning machine can out-test, out-learn and out-grow teams a hundred times its size.
We hold ourselves to the standards of Apple, Spotify and MasterClass. While others hide their numbers, we let our results speak.
YOUR MISSION
As Head of Growth, you’ll be the architect of a growth machine that learns by itself. In a classic growth team, people do the work and AI helps. Here it’s the other way round: agents do the work, you design the system, set the bar and make the calls.
You don’t just manage a team, you get your hands dirty to identify growth levers and pull them yourself.
YOUR 5 MAIN MISSIONS
1. Build the agents that run growth
The old way: a Head of Growth hires people and agencies, briefs them, and checks their work.
At Paradox: your team is mostly agents, and you build them yourself. You write their skills, plug them into our tools (ad accounts, CRM, webinars, email, analytics) and write the evals that tell an agent what "good" means: a hook that stops the scroll, a page that converts, a test that’s worth its cost.
You give autonomy step by step: first an agent suggests, then it acts once you approve, then it acts alone within a budget and guardrails. You move each agent up a level only when its track record earns it.
Example: in your first month you build the agent that reviews every ad set overnight. At first it only suggests. By month three it pauses losers and moves budget by itself within limits you set, and you only see what was decided and why.
2. Build an acquisition engine that runs day and night
The old way: a media buyer checks the dashboards each morning, changes bids by hand and waits for the weekly report.
At Paradox: you build and improve the agents that watch every ad set, page, webinar and email sequence around the clock. They spot what’s dropping, find the cause, fix it or launch a variant, and report what they did.
You give autonomy step by step: an agent can pause an ad on its own long before it can raise a budget. You decide where each agent stops and a human takes over.
Example: show-up rate drops on Tuesday’s webinar. Before anyone notices, an agent has found the cause (a reminder that never went out), fixed it and started testing a new reminder sequence.
3. Build a creative machine with taste
The old way: a creative team briefs an agency, waits two weeks and gets five ads.
At Paradox: you build the agents that turn every live, interview and member story into hooks, ad cuts and shorts, that study what the best advertisers in the world are running, and that write in our voice.
Your taste becomes the eval: you write down what makes a great hook, a great cut and a great story, in rules and examples the agents check themselves against. Each time you reject something, the reason goes back into the rules.
Example: one webinar becomes dozens of hook variants by the next morning. You keep the best few, and next week the agents’ first drafts are already closer to your bar.
4. Build measurement that checks itself
The old way: an analyst rebuilds the numbers in a spreadsheet at the end of the month, and nobody trusts them.
At Paradox: you build the agents that connect every touch to cash, from ad to lead, live, sale and transformed member, and that audit the tracking after every campaign.
No number, no test: you define the KPIs that matter (CAC, show-up, close rate, LTV, payback), and agents refuse to launch a test that isn’t wired to one.
Example: after each live, an agent cross-checks the ad platform, the CRM and our analytics, finds where they disagree and fixes the leak before the next live.
5. Make the machine better every week, and keep the soul
The old way: lessons live in people’s heads and leave with them.
At Paradox: every result goes back into the company brain, and agents read it before proposing anything. The machine never learns the same lesson twice.
You measure the machine itself: tests per week, time from idea to live, the share of the work done without a human, and how many agents moved up an autonomy level this month. You push those numbers up.
And always: no fake scarcity, no fake claims, real stories told with consent. The machine can scale, but it doesn’t decide what’s right. You do.
How you’ll know you’re AI-native: you spend your time building agents, writing evals and making calls, not doing the work by hand. Most experiments are generated and shipped by agents. The output grows every month while the headcount doesn’t.
You’ll own these KPIs: LTV / CAC ratio, funnel conversion rate by stage, experiments launched per week.
Desired outcomes: conversion rates 2x better than baseline, growth experimentation that runs continuously, a predictable growth engine.
Tools you’ll master: GA4, Optibase, Vidalytics, Clarity, Hotjar, Amplitude, Figma, Looker Studio, VWO, Optimizely, Claude Code, and whatever helps you see what others miss.
THE IDEAL AI NATI