The AI Academy
All 12 plans
08IntermediateDo it the narrow way

AI UGC and Product Creative

Produce the product visuals and ad creatives that e-commerce brands need every week.

Creative studio

Briefs in, approved assets out

Active projects
6
Assets rendered
135
Awaiting review
6
Invoiced
€3,150
Studio pipeline
Brief
  • Nordvik · summer range
  • Havenzicht · autumn
Generating
  • 6 SKUs on white
  • 3 lifestyle scenes
Human review
  • 6 SKUs on white
  • 2 shots · hands and text
Delivered
  • 22 assets
Rendering now
  • Nordvik · 6 SKUs on white84%
    Product shots
  • Nordvik · 3 lifestyle scenes47%
    Styled set

Example figures, shown to illustrate how the workflow behaves.

Open full screen · recording mode — sized for vertical video.

Read this before the opportunity

This one demonstrates far better than it performs. Its content score is 8.6 against a business score of 6.2 — a gap of 2.5 points. That pattern is exactly what the get-rich-with-AI industry is built on: strategies that look spectacular in a short video and disappoint in practice.

The before and after is the entire pitch, and it lands in under two seconds.

Business score
6.15
Content score
8.60
Combined
7.58

Follow these prompts to start building your income stream

6 steps. Paste each prompt into Claude, use what comes out, move to the next. The last one ends in something live — not in a plan to make something live.

6 steps
  1. 01

    Make three sample ads

    Free sample

    One product category, and three finished ads for a brand you do not work for.

    In this step you make the thing that gets you hired. Brands do not read pitches from new creative suppliers, they look at work — so you make work for a brand you admire, unpaid and unasked. You end with a sample pack that does the selling in every message you send afterwards.

    1 — Choose the category
    I want to sell AI-produced ad creative to brands. Help me choose one product category.
    
    Categories I understand as a customer: [YOUR CATEGORIES]
    What I can produce today: [YOUR SKILLS]
    
    For each category:
    1. What the ads in this category have to prove before someone buys.
    2. Which formats dominate — demonstration, testimonial-style, before and after, unboxing — and why that one.
    3. What AI-produced creative does well here, and where it visibly falls apart.
    4. Which claims are regulated in this category, and what that means for what I can show.
    5. How brands in this category usually buy creative, and what budget they treat as small.
    
    Then rank by "a small brand could hire a newcomer here this month", and explain the top one.
    2 — Build the sample pack
    I am making an unpaid sample pack for [BRAND], which sells [PRODUCT] in [CATEGORY].
    
    Their current ads, as far as I can see: [WHAT THEY RUN NOW]
    What their product genuinely does: [PRODUCT REALITY]
    
    Give me three complete ad concepts, each testing a different angle. Per concept:
    1. The hook, in the first two seconds.
    2. The full script or on-screen copy.
    3. The shot list, with what is generated and what must be real product footage.
    4. The generation prompts for the parts I will produce with AI.
    5. What this concept is testing that the others are not.
    6. The claim boundary: what this ad must not say in this category.
    
    Stay inside what the product actually does. A sample that overclaims tells the brand exactly the wrong thing about working with me.

    Claude for concepts and scripts; your image and video tools for production.

    [YOUR CATEGORIES]
    Products you buy and understand. You need to know what makes a buyer hesitate.
    [BRAND]
    A real brand whose ads you think could be better. Pick one small enough to reply to you.
    [PRODUCT REALITY]
    What it actually does, from their own site and reviews. Never invent a benefit in a sample.

    Then: Finish all three. An unfinished sample pack is a pitch again, and pitches get ignored.

    Where the prompt stops: Using a brand's product in unsolicited sample work is normal in this trade, but do not publish it as if they hired you, and do not use their logo in a way that suggests endorsement.

  2. 02

    Put your offer online

    In the package

    A page with the samples, what you deliver, and what it costs.

    In this step the samples get a home and a price. Creative services sell badly when they are priced per hour and well when they are priced per deliverable, because brands buy a quantity of ads, not your afternoon. You end with something you can link in one line.

  3. 03

    Send it to fifteen brands

    In the package

    Fifteen named brands approached with work attached.

    In this step you use the samples as the opener. The message is short because the work is the message: here are three ads I made for you, here is what each one tests, here is what a batch costs. You end with conversations rather than a portfolio nobody saw.

  4. 04

    Agree the job, take a deposit

    In the package

    A brief both sides signed, a deposit in, and a delivery date.

    In this step a reply turns into paid work. Creative jobs go wrong in the same two ways every time: the brief was verbal, and the money arrived last. Settling both in one short document is what makes the first job profitable instead of educational.

  5. 05

    Make the first paid batch

    In the package

    A delivered batch with the variants a buyer can actually test.

    In this step someone pays you and you deliver work that is easy to run. What separates a useful creative supplier from an expensive one is variants: the same concept in the formats and lengths the platform needs, named so a media buyer can find them.

  6. 06

    Report back and get the next job

    In the package

    A short report they can act on, and a second batch agreed.

    In this step you do what most creative suppliers skip: ask what happened. Media buyers rarely volunteer results, but they answer a direct question — and a supplier who asks becomes the one who iterates instead of the one who guesses. This is also the conversation where a retainer starts.

5 of the 6 steps are in AI Money

The sample above is a real step, not a teaser — it is the same depth as the rest. The package adds the other 5 for this strategy, and the same treatment for the other eleven.

Your first 7 days

  1. 1Pick one product category and make six ad variations for a brand you do not work for yet.
  2. 2Label every asset as AI-generated in the file name and the delivery note.
  3. 3Write the one-page offer: how many assets, how fast, at what price.
  4. 4Send the sample pack to ten brands that are already running ads.

Your first 30 days

  1. 1Deliver a first paid batch with a disclosure line the brand can copy into its ad.
  2. 2Report which variation performed and why, in the brand's own metrics.
  3. 3Convert the best-performing client into a monthly batch agreement.
  4. 4Build a reusable hook library from what actually worked.

What you need

core means you cannot skip it. The rest only if you need it.

  • HeyGencorefree planAvatar presenters for product spots
  • Runwaycorefree planMotion and video generation
  • Kling AIfree planProduct motion shots
  • Photoroomcorefree planProduct cutouts and scene placement
  • CapCutfree planEditing and caption styling
  • ElevenLabsfree planVoice-over in multiple languages
  • Recraftfree planBrand-consistent still creative
Want the research?— what it is, what it costs, and how we scored it

What you can actually watch

A flat product photo becoming a styled shot, then a model shot, then an ad.

What it is

You produce the product visuals and ad creatives that e-commerce brands burn through every week — hooks, variations, avatar-presented spots — using generative tools instead of a studio.

Why this opportunity

The production lever is real and the buyers exist: DTC brands and performance agencies need new creative constantly, and tooling is cheap — Creatify Starter is $39 a month and Pro $99 (Tier 1, creatify.ai, LAST VERIFIED: 2026-09-07). Visual demo scored 10 of 10: the before-and-after is the entire pitch.

Who this is for

For someone with an eye for what stops a scroll, who is willing to disclose AI use openly. A bad fit for anyone who wants to pass synthetic footage off as authentic customer content — that is precisely where the legal and commercial risk sits.

The business model

Per video or in packages, sometimes on a monthly creative retainer. Retainers are the only version with defensible economics, because single videos are where price competition is hardest.

How money is made

A fee per asset or per package. Our research found average UGC rates fell roughly 44% between 2024 and 2025 while the creator pool grew 93% (Tier 4 aggregation, UNCERTAIN on the exact figure, but the direction is consistent with every other dataset).

Who pays

DTC brands and performance agencies that already spend on ads — people with a measurable reason to buy creative and a budget line for it.

Finding customers

Marketplaces, cold outreach and portfolio. Your portfolio is the pitch; nobody reads a proposal for creative work.

Delivery

Hook, avatar or product render, then variations for testing. Volume is the deliverable — brands do not want one video, they want twelve to test.

Where AI actually helps

Very high in production and variation. Zero in the thing that makes an ad work, which is knowing what to say first. That gap is why the commodity end of this market is being priced away.

Skills you need

Low to medium technically; the real skill is direct-response judgement — hooks, pacing, and what an offer has to prove in three seconds.

What it costs to start

Low: a generation subscription and an editor. Creatify Starter $39/month, Pro $99/month (Tier 1, creatify.ai). Avatar tools sit in the same band.

How to price it

Package pricing per batch of variations, or a monthly retainer for an agreed output. Per-video pricing invites a race to the bottom that you will lose to someone with cheaper time.

Common mistakes

Hiding the AI. A 2026 Harris Poll / 4A's / Infillion study found 78% say AI makes ads feel less authentic, 73% trust an ad less when they suspect AI, and 63% are less likely to buy from a brand using AI-generated ads (Tier 2). Disclosure is also becoming law: New York legislation in force since 9 June 2026 requires ads using AI-generated people to say so, with fines of $1,000–$5,000, and the obligation sits on the merchant you produce for. The FTC's rule against fake reviews and testimonials (final 14 August 2024, effective 21 October 2024, penalties above $50,000 per violation, Tier 1) covers simulated endorsements — it acted against Rytr, an AI review generator, in December 2024.

Scaling

Scalability 6, automation 9, competition 10 — the most crowded scores in the set. Scale comes from retainers and from owning a niche's visual language, not from producing more videos per hour.

The version that holds up

Sell the testing system rather than the assets: a monthly batch, a naming convention, and a report on which hook won. Tellingly, human-made UGC is now priced as premium — six human-made videos for $1,000 (Tier 1, icon.com) — so the durable position is directing AI openly, not competing with it on price.

Time to first revenue

1–6 weeks

1–6 weeks to a first paid job through marketplaces, at rates that are falling. Speed to first revenue is the strong point here; durability of that revenue is the weak one.

How it scored

Ten axes, scored. Risk scored 6 out of 10, where higher is worse.

Content potential9/10Editorial assessmentCuriosity8/10Editorial assessmentVisual demo10/10Editorial assessmentMoney visualisation7/10Editorial assessmentMarket demand7/10Editorial assessmentSimplicity6/10Editorial assessmentScalability6/10Editorial assessmentTool ecosystem8/10Editorial assessment

The state of this market

Consumer research points the other way on AI advertising: a 2026 Harris Poll study found 78% say AI makes ads feel less authentic and 63% are less likely to buy from a brand using AI-generated ads. That does not make the service unsellable — brands are buying it — but it means the winners will be the ones disclosing openly rather than hiding it. Disclosure is also becoming law rather than a choice: New York legislation in force since 9 June 2026 requires advertisements using AI-generated people to say so, with fines of $1,000 to $5,000 and the obligation sitting on the merchant you are producing for.

Not sure this one fits you?

Answer a few questions about budget, time and skills and see which of the eighteen match your situation — and which to avoid.

Run the Opportunity Finder