neuro8.aiSign in

Less about what worked. More about what works next.

neuro8 pulls your live Meta ads daily, scores them on brain activation across six regions, and tells you why one ad is working and what to ship next.

Beta opens May 31. Free for everyone on the list.

  • Built on Meta's TRIBE v2
  • Algonauts 2025 winner
  • Used at V8 Media on $2B+ in ad spend
A brain analysis showing 6 brain regions scored on a real Meta ad. Memory region is flagged as critical.

You're not running ads. You're running Whac-A-Mole.

Every operator we know lives the same week.

  • My winning ad just stopped. I have no idea why.

  • I'm burning cash to learn what doesn't work.

  • This one's working. I just don't know what to ship next.

You're swimming in a sea of sameness. Tool bloat keeps growing. Dashboards keep multiplying. And the why still lives in the brain.

Three steps. No new tabs. No new tool.

Connect once. Wake up to the why every morning.

  1. Step 01

    Connect your Meta workspace.

    One click OAuth. We see what your account sees. Nothing more, nothing less.

  2. Step 02

    We pull your live ads every day.

    Each ad gets scored on six brain regions. Emotion. Desire. Logic. Memory. Trust. Language. Your account, not a generic benchmark.

  3. Step 03

    You see why, and what to ship next.

    Each ad gets a tag. Cut. Optimise. Rescue. Scale. Fatigue. With a one-line reason and a concept for the next test.

This is what your ad looks like to a brain.

Six regions. One ad. One reason it's winning.

A six-region brain heatmap rendered next to a real Meta ad creative. Emotion at 0.84, Memory at 0.79.
  • Emotion0.84
  • Desire0.71
  • Logic0.42
  • Memory0.79
  • Trust0.66
  • Language0.58

This ad lights up Emotion and Memory. That's why it's working. The next test should keep the emotional hook and lift Logic with a clearer offer.

The why lives in the brain. If you can't see the why, you can't predict what works next. This is the readout most tools never show you.

Real fMRI. Real model. Real benchmark.

This isn't a vibe. It's a peer-reviewed brain model from Meta.

  • 1,117
    hours

    Of real fMRI scans from 720 people. The training data behind TRIBE v2.

  • 263
    teams beaten

    TRIBE v2 won the Algonauts 2025 challenge against 263 research teams.

  • ~82%
    directional

    Accuracy on the Bercea 2017 benchmark. Peer-reviewed, not vendor-claimed.

  • Open
    source

    TRIBE v2 lives on HuggingFace under facebook/tribev2. You can verify it yourself.

We didn't invent the brain model. Meta did. We built the agency layer on top of it.

Why neuro8 isn't another Motion or AdCreative.

Most tools tell you what worked. neuro8 tells you why, and what works next.

neuro8

Signal type
Brain activation, six regions
What it predicts
Why an ad works, what to ship next
When it predicts
Before you spend, daily
Account-specific?
Yes, your live ads daily
Status
Live

Motion

Signal type
Performance metrics (ROAS, CTR)
What it predicts
What already worked
When it predicts
After spend, after the fact
Account-specific?
Yes, your data
Status
Live

AdCreative.ai

Signal type
Generic creative scoring
What it predicts
A predicted score on a stock benchmark
When it predicts
Before you spend
Account-specific?
No, generic model
Status
Live

Pattern89

Signal type
Visual pattern matching
What it predicts
Visual element correlations
When it predicts
Before you spend
Account-specific?
No, generic model
Status
Defunct
Jandre de Beer, founder of neuro8.ai and V8 Media.
Why we built this

Built by an operator who got tired of guessing.

I run V8 Media. Over two billion in ad spend across 600 plus clients. We tested every major AI ad tool on the market for our agency. None of them answered the question we actually had. Why one ad fires and the next one stops. So we built our own.

For ten years I watched the same pattern. A winner stops winning. Nobody can tell you why. So you ship ten more variations and pray.

Now we score every ad on the same brain regions a $20,000 fMRI study would. Every morning. Inside our own dashboard. The team stopped guessing within a week.

If you run ads for a living, this is the tool I wish I'd had ten years ago.

Jandre de Beer · Founder, V8 Media & neuro8.ai

Every prediction sharpens the next one.

From feedback loop to flywheel.

  1. 1

    Predict

    Score your live ad on six brain regions before it spends.

  2. 2

    Ship

    Launch the ad in Meta.

  3. 3

    Measure

    Real CTR, ROAS, and watch-time come back.

  4. 4

    Learn

    Every prediction-to-outcome pair sharpens the model.

This is the world's first neural-to-CTR map. Every brand on neuro8 makes the next prediction smarter. Joining the beta means your data shapes the model, not the other way around.

What you're probably thinking right now.

  • Is this just another AI tool?
    No. Most AI tools tell you what worked. neuro8 tells you why, and what works next. The model is from Meta, trained on real brain scans. It scores your live ads, not a stock benchmark.
  • Neuroscience in marketing feels like woo. Didn’t Pattern89 fail?
    Fair. Pattern89 used visual pattern matching, not brain data. neuro8 runs on TRIBE v2, an open-source brain model trained on 1,117 hours of real fMRI. You can verify the paper and the model on HuggingFace.
  • Isn’t 720 subjects a small sample?
    1,117 hours of fMRI from 720 people is one of the largest open neuroimaging datasets ever released. TRIBE v2 won the Algonauts 2025 challenge against 263 research teams on it.
  • Sounds expensive for an uncertain ROI.
    Beta is free. After beta we’ll charge a flat monthly fee, less than one bad ad-set test. If it doesn’t pay for itself in saved spend, cancel.
  • I don’t trust black-box AI.
    You see the six region scores for every ad, every day. You see the reason. You see what to test next. Nothing is hidden behind a single magic number.
  • Isn’t this just analytics like Motion?
    Motion is a reporting tool. It tells you what already happened. neuro8 predicts what will happen, on your account, before you spend.
  • What does the beta cost?
    Nothing. Beta opens May 31. Everyone on the waitlist gets in free.
  • What happens after beta?
    Paid plans launch later in the year. Beta members lock in founder pricing for 12 months.

Less about what worked. More about what works next.

Beta opens May 31. Free for everyone on the list.

No spam. No drip. One email when the beta opens. That's it.