Bulletin · #1 · 15 May 2026
The Neuromarketing Lab: the method, and the ECLAG case
How Pianeta.Studio validates communication concepts with two complementary AI technologies — Swarm and Neural Prediction — while keeping the decision firmly in human hands.
Who this is for
Some communication decisions matter more than others. A new launch. A campaign on a sensitive topic. A repositioning. These are the moments when a wrong brief costs months — and when the creative team’s instinct, even when right, struggles to find a language to defend itself in front of the client.
Pianeta.Studio’s Neuromarketing Lab exists for those moments. It isn’t an off-the-shelf service: it’s a research and development capability that applies two complementary AI validation technologies to creative work, before choices are locked in.
We don’t replace creativity, and we don’t steer it. We work with organizations that recognize the scientific method as a posture — not as a guarantee of results, but as an honest way to interrogate a hypothesis.
The principle: AI safety and human in the loop
Before describing the tools, it’s worth saying what we don’t do. The lab rests on three non-negotiable principles, written down in every project.
01. AI is a stimulus, not a decision
Models produce evaluation data — scores, predictions, comparisons across variants. They don’t pick the concept. They don’t sign off on the brief. They return evidence, not verdicts.
02. Human in the loop
The client knows their sector. The creative team knows the craft. AI is one more lens, never a substitute. It’s a condition stated in writing in every project, not an implicit premise.
03. Transparent about limits
The models have stated limits: the samples they were trained on, the brain regions they cover, the scale at which they predict. We share them with the client at the start, not at the end — when they can still calibrate expectations.
In practice: tests inform the decision table, we respond creatively, we re-test. Sometimes we go our own way knowing we’re off the data — and that’s a legitimate, declared choice.
The method: two complementary tools
The lab works with two technologies that interrogate different objects of the creative process. They don’t overlap: they complete each other.
Swarm Analysis (strategic phase)
Pianeta MiroFish
Validates concepts, positioning, messaging, and briefs. 50–80 AI agents calibrated on real buyer personas interrogate the proposal with iterative feedback and anticipated objections. Used before production, when the cost of a course correction is still low.
Output: co-validated brief · iterative score · objections list
Neural Prediction (evaluation phase)
TRIBE v2 · Meta AI Research
Validates audio, video, voice, and music stimuli. A foundation model predicts fMRI activity across 20,484 cortical points, without real subjects. Used to choose between variants when the concept is set and the decision is one of execution.
Output: predictions across 6 brain regions · comparative score
The Swarm interrogates the simulated market on the concept; Neural Prediction interrogates the simulated brain on the stimulus. Together they cover the full decision cycle.
Recent case: ECLAG “Choose to See Them”
The most honest way to explain a method is to show what it did on a real case. On the ECLAG Choose to See Them campaign — a sensitive topic, European distribution — the lab intervened in three distinct phases.
Phase 1: Pre-production
Swarm analysis on two competing creative proposals. 32 AI agents distributed across 5 demographic groups. The winning proposal earned 30 votes out of 32 — a clear margin, not a coin toss. The key insight came on the “uncomfortable” claim: what the creative team had intuited as a strength, the Swarm quantified as such, isolating where the discomfort worked and where it risked pushing people away.
Phase 2: Full-campaign validation
Pianeta Validation Engine v1. 29 agents generated from a knowledge graph of 116 nodes, 72 rounds of questioning, 312 pieces of content generated and evaluated. Not a single measurement, but a reaction map across the entire campaign before final production.
Phase 3: Final stimuli
Neural Prediction with Meta AI Research’s TRIBE v2. 12 predicted fMRI evaluations on 20,484 cortical points to choose between the final stimulus combinations. From here came the final combination recommended at the decision table.
What the method did: in all three phases the method confirmed and quantified the creative team’s intuitions. It never replaced the client’s decisions.
The campaign is in European distribution and live online: choosetoseethem.childsafetyineurope.com
The full AI-augmented pipeline
The two technologies aren’t isolated events: they fit into a six-phase pipeline that runs across the project from kickoff to delivery.
- Discovery Sprint — Framing the problem, objectives, constraints, audience.
- Swarm Analysis — Concept / positioning / messaging validation on the simulated market.
- Creative production — Developing the stimuli on the co-validated brief.
- Neural Prediction — Choosing between variants based on cortical prediction.
- Real Focus Group (optional) — Verification with real people when the case requires it.
- Final production — Execution and delivery.
Phases 1–4 are typically enough for a mid-scale project. Phases 5 and 6 are modular: triggered when the case calls for them, not by default.
The lab team
Massimiliano Mauro — Strategic direction · AI safety / human-in-the-loop
Sude Eroglu — Behavioral science, neuromarketing and AI safety researcher · methodological validation and ethical framing
Fabrizio Ciampini — DevOps and R&D infrastructure · TRIBE v2 / MiroFish integration
This is the method. If you have a communication decision that matters — a launch, a campaign on a sensitive topic, a repositioning — and want to find out whether the lab can be useful for your case, write to max@pianeta.studio.
Prepared by Pianeta.Studio · May 2026 · Bulletin editorial document
