INM Neuro LAB
ESENPT
Let’s talk
IN DEVELOPMENT · ASSISTED PILOTS
NEUROPERFORMANCE

What people feel,
connected to what media delivers.

Applied neuroscience and performance marketing still live apart: an asset is tested, recommendations are delivered, and no one learns how it did in-flight. NeuroPerformance closes that loop.

Let's talk about a pilotSee NeuroLab Core
THE GAP

Two worlds that don't talk.

Neuromarketing

Measures attention, cognitive load and emotion, but rarely revisits the asset after launch.

Performance

Measures CTR, CPA and ROAS, but can't explain why a creative works or wears out.

NeuroPerformance joins both readings in a single record per creative.

CLOSED LOOP

Every campaign improves the next.

Each creative that is evaluated and then run adds a data point linking neuroscience signal and performance. With more cases, our own models get sharper.

  1. 01

    Neural evaluation

    Attention, cognitive load and predicted engagement for each asset before spend.

  2. 02

    Media launch

    The creative runs on Meta, TikTok, Google or programmatic with its neuro profile attached.

  3. 03

    Real outcome

    Campaign CTR, CPC, conversion rate, CPA and ROAS are brought in.

  4. 04

    Proprietary learning

    The cross feeds models that recommend what to fix and what to scale next.

OWN R&D

Neural prediction. Proprietary INM models.

We start from attention prediction and build proprietary models that cross three data layers. That is where the innovation lies.

Creative DNA

Neuroscience signals and attributes of each asset.

Performance Data

Real media results per creative.

Market Context

Category, platform, format and local calibration.

CORE MODULES · 7 tools

Creative intelligence engine.

Seven modules that take neural prediction from the creative table to paid media: filter, invest, rotate, reallocate, generate, protect brand DNA and test in real context.

01

Neural Creative Pre-filter

Which of these variants deserve budget?

The problem
Teams produce 5 to 30 variants before launch, and most are discarded after spending real budget on A/B tests.
How it solves it
Every image or video goes through neural prediction and returns four metrics in seconds: Focus (where people look), Clarity (how easy it is to process), Cognitive Demand (mental effort) and Engagement (emotional activation), plus a heatmap of what gets seen and what stays invisible.
What your team gains
Out of 10 variants, you spot the best before spending on media, and A/B testing runs only among pre-validated winners.
Flow
  1. Upload the asset
  2. Neural prediction
  3. Scores + heatmap
  4. Variant ranking
  5. Run only winners
FocusClarityCognitive DemandEngagement
02

Neuro-Programmatic Bidding

Why pay the same for a brilliant asset and a mediocre one?

The problem
In programmatic, the bid is identical for every creative in a campaign. The algorithm takes days or weeks, and budget, to learn which one works.
How it solves it
When an asset combines high Focus and low Cognitive Demand, the mix that tends to anticipate better conversions, it gets a more aggressive bid. Low scores lower the bid or exclude the asset. It works like an algorithmic attention trader, with a daily spend cap as a safety net.
What your team gains
More return per budget and a much shorter learning period on Meta and Google.

Illustrative exampleFocus > 70 and Engagement > 65 → bid 1.5×. Focus < 40 or Cognitive Demand > 80 → exclude.

Flow
  1. Neural score
  2. Client rules
  3. Bid 1.5× / 1× / exclude
  4. Run on Meta and Google
  5. Real CTR, CPC, ROAS
  6. Back to the model
FocusCognitive DemandBidDaily cap
03

Creative Decay Predictor

When will this asset wear out?

The problem
Creative fatigue is detected after it happens: CTR dropped this week, and only then are new assets requested.
How it solves it
It tracks both how your asset holds attention in its real feed and the rising quality of competitor ads in your category. If the market improves while your asset stays static, it alerts you ahead of time.
What your team gains
Creative teams work proactively, and dead periods paying for ads that no longer work disappear.

Illustrative example"This asset may lose relevance in about 7 days. Prepare replacements."

Flow
  1. Weekly feed capture
  2. Your asset analysed
  3. Competitors analysed
  4. You vs. market
  5. Estimated wear-out date
  6. Alert + suggested variants
Attention over timeCompetitionCTR drop
04

Attention Arbitrage Engine

Where does each dollar buy the most attention?

The problem
The same asset does not capture the same attention on every channel, and the cost of showing it varies widely.
How it solves it
It crosses the attention an asset generates on each channel with what it costs to show it there, and computes a proprietary efficiency indicator: Attention per Dollar (APD). Like a financial arbitrageur, it finds where attention is undervalued.
What your team gains
Concrete reallocation recommendations to get more total attention from the same budget.

Illustrative example"Your asset delivers 3× more attention per dollar on TikTok than on LinkedIn: shifting budget raises total attention without spending more."

APD formula and calculationAPD = (Focus × 0.6 + Engagement × 0.4) × Reach ÷ CPM

Attention per Dollar: how much attention each budget unit buys on each channel.

ChannelScoreCPMEst. reachAPDRecommendation
Instagram Feed72%$8.5045K8.47Keep
TikTok Feed81%$5.2062K15.58Increase +40%
YouTube Pre-roll65%$12.0038K5.42Reduce
LinkedIn Feed58%$15.0012K3.87Shift to others
Illustrative values to explain the calculation; not a real client.
Flow
  1. Asset analysed once
  2. Real or benchmark CPM
  3. APD per channel
  4. Channel ranking
  5. Suggested reallocation
FocusEngagementReachCPM
05

Generative Creative Loop

What if variants were designed with data from the start?

The problem
Producing variants takes days of design, review and edits, and many are designed on intuition rather than data.
How it solves it
Three coordinated agents: one generates variants from the brief, another scores them with neural prediction and drops the weak ones, and a third studies what winners share (colours, CTA position, composition) and issues improvement instructions. The loop runs three times.
What your team gains
A week-long process compressed into one session, with the best variants going to a human panel for the final call.
Flow
  1. Client brief
  2. Designer agent
  3. Evaluator agent
  4. Strategist agent
  5. Repeat ×3
  6. Top 3 to expert review
  7. Launch
VariantsAttentionWinning patterns
06

Digital Creative Twin

Which visual pattern really works for your brand?

The problem
Every brand has colour, composition and style combinations that work, but that knowledge lives in a creative director’s intuition, not in a system.
How it solves it
Each asset is broken down into its visual DNA: dominant colour, saturation, faces, CTA position, text density, photographic style. With real results, the platform builds a profile of what works for your brand and flags new assets that drift from the winning pattern.
What your team gains
A strategic asset exclusive to your account that sharpens with every piece. It doesn’t replace creativity: it informs it.

Illustrative example"You can innovate, but when you drifted towards cool tones, performance dropped."

Flow
  1. New asset
  2. Visual DNA extraction
  3. Compare with your twin
  4. Drift detected
  5. Evidence-based alert
ColourFacesCTAText density
07

In-Context Simulation + Voice of Attention

How does your asset look where it actually competes?

The problem
An ad that looks great in isolation can disappear inside a feed, a web page or a crowded shelf.
How it solves it
It places the asset in a realistic mock-up of its final environment and measures how much attention it captures against everything around it. Voice of Attention adds audio: it analyses vocal tone and energy second by second and flags when image and voice contradict each other.
What your team gains
Decisions grounded in real exposure context, and videos where what is seen and heard push the same way.
Flow
  1. Asset
  2. Feed, web or shelf mock-up
  3. Full-scene analysis
  4. In-context attention delta
  5. Audio + image cross
Contextual attentionVoice sentimentDissonance
HYBRID LAYER · AI + EXPERT · 3 tools

Where technology meets judgement.

AI analyses thousands of assets an hour; contextual judgement is still human. These modules combine both to produce insights no AI can deliver alone.

H-01

Expert Review Panel

What can’t the machine see?

The problem
AI doesn’t know if a brand just repositioned, if there is a crisis underway, or if a high-scoring tone is culturally wrong for a market.
How it solves it
Specialists in applied neuroscience, creative direction and media strategy review automated analyses and add what the machine can’t: fit with the brand moment, audience appropriateness and perception risks. The system tracks each reviewer’s historical accuracy.
What your team gains
A report combining data and expert interpretation, with the most accurate reviewers on the most critical accounts.
Flow
  1. Automatic review package
  2. Expert assigned
  3. Brand and cultural judgement
  4. AI + human report
Brand contextCultural fitRisk
H-02

Creative War Room

What if optimisation happened live?

The problem
Conventional optimisation bounces around by email: each round of proposals and feedback takes one to three days.
How it solves it
A live session where the INM team uploads variants, shows scores instantly and edits in real time. Your team sees how moving a CTA, changing a colour or resizing an element directly affects attention.
What your team gains
Optimisation shifts from opinion to data, fast and transparent, with a full log of every iteration.
Flow
  1. Session booked
  2. Variants on screen
  3. Live edit
  4. Instant score
  5. Session log
Before / afterAttentionIterations
H-03

Performance Feedback Loop

Does a Focus of 72 translate into sales?

The problem
An attention score doesn’t say whether it means a 2% or a 4% CTR. Only real campaign data can answer that.
How it solves it
Once ad accounts are connected, the platform records what happens to each asset after launch: clicks, cost and return. It builds what neither media platforms nor attention studies hold separately: the intersection of neural prediction and real performance.
What your team gains
A calibration view showing how attention and results relate in your account, and the fuel for every proprietary model.
Flow
  1. Connect your accounts
  2. Results per asset
  3. Cross with neuro scores
  4. Focus vs. CTR calibration
  5. Sharper models
CTRCPCROASConversions
PROPRIETARY MODELS · 5 tools

Neuroscience predicts attention. INM predicts business results.

Five models that cross neuroscience signals, visual DNA and market context. That gap is where intellectual property is built.

ML-01

Performance Predictor

How much will this asset sell?

The problem
Knowing where people look isn’t enough: a CFO needs business results.
How it solves it
It estimates the expected CTR or ROAS of an asset before launch, by crossing three layers only this platform accumulates: neuro signals, visual DNA and market context.
What your team gains
Neuroscience translated into metrics finance understands. Accuracy grows with every real campaign.

Illustrative example"If I run this asset on Instagram, what CTR can I expect versus the previous one?"

Flow
  1. Neuro signals
  2. Visual DNA
  3. Market context
  4. Expected CTR / ROAS range
Expected CTRExpected ROAS
ML-02

Creative Recommender

What worked for brands like yours?

The problem
A brand new to the platform has no history of its own for recommendations.
How it solves it
Like Netflix or Spotify recommenders, it finds comparable brands by category, country and audience, identifies the visual patterns of their winning assets and prescribes edits to yours.
What your team gains
Useful recommendations from day one, increasingly specific as you build your own data.

Illustrative example"Top-ROAS fintech brands in LATAM use a human face and an action-verb CTA. Your asset meets 2 of 5 criteria."

Flow
  1. Your asset
  2. Comparable brands
  3. High-performing assets
  4. Winning criteria
  5. How many you meet
CategoryCountryPatterns
ML-03

Anomaly Detector

Which brilliant exception are we missing?

The problem
A model that only learns normal patterns misses the assets that break the rules and still work.
How it solves it
It looks for assets that defy convention and, against expectations, deliver exceptional results. It analyses why they work and turns them into hypotheses to test in future campaigns.
What your team gains
Opportunities a conservative approach would never find, such as categories where visual complexity sparks curiosity.
Flow
  1. Normal distribution
  2. Outlier detected
  3. Positive anomaly
  4. Hypothesis
  5. Controlled experiment
OutliersHypotheses
ML-04

Attention Decay Forecaster

On what exact day will it start to drop?

The problem
Spotting wear-out as it happens leaves little room to react.
How it solves it
It learns from the historical trajectory of many assets and competitive context to forecast the date your asset will lose effectiveness, with a confidence interval.
What your team gains
Weeks of lead time to prepare replacements instead of reacting in a rush; near the threshold, it can trigger the generative loop.
Flow
  1. Historical series
  2. Competitive context
  3. Forecast with interval
  4. 7-day alert
  5. Replacements ready
TrendSeasonalityConfidence interval
ML-05

Causal Impact Analyzer

Did the creative change cause the lift, or something else?

The problem
Correlation isn’t causation: a better CTR may come from a better audience or the season, not the creative change.
How it solves it
It isolates each variable’s effect with statistical techniques and estimates how much of an improvement comes from the creative change versus other factors.
What your team gains
Creative optimisation moves from art to quantifiable science, with causal relationships by category.
Flow
  1. Creative change
  2. External factors controlled
  3. Isolated effect
  4. Quantified causal link
Causal effectSeasonalityAudience

Models are in development and their accuracy depends on the volume of real cases. They are not presented as guaranteed results.

COMPLEMENTARY SERVICES · 6 tools

An ecosystem that feeds the core.

Each service solves a different client problem and adds data to the platform core.

S-01

Brand Consistency Engine

Do all your offices and agencies speak the same visual language?

The problem
Teams across countries and agencies produce separately, each with its own palette, and every drift from the winning pattern costs performance.
How it solves it
An automatic guardian checks every asset against the brand’s digital twin and a rulebook (colour ranges, minimum Focus, logo position, typography). It works as a mandatory approval gate or as an informative alert.
What your team gains
Consistent visual identity no matter who produces the content.
Flow
  1. Asset from any office
  2. Compare with the twin
  3. Brand rules
  4. Approved / fix
ColourLogoTypographyMinimum Focus
S-02

Influencer Creative Intelligence

How much attention does your product get inside the creator’s content?

The problem
Influencer selection relies on vanity metrics: followers, likes and engagement rate. None answers what matters.
How it solves it
It analyses each candidate’s latest posts and measures how much attention the product gets within their content. A smaller creator can give the product more visibility than a much bigger one.
What your team gains
Candidates compared by product attention per dollar invested, not by followers.
Flow
  1. Candidates
  2. Latest posts analysed
  3. Product visibility
  4. Attention ÷ cost
  5. Recommendation
Brand visibilityAttentionCost per post
S-03

Neuro-Email Optimizer

Is your CTA seen before the email goes out?

The problem
Emails are optimised with open and click rates that are only known after sending.
How it solves it
It analyses each email as a full image and shows whether the main CTA gets enough attention, whether the hero image distracts, or whether the header carries too much cognitive load.
What your team gains
Actionable recommendations before reaching a single subscriber.

Illustrative example"Your CTA gets only a fraction of attention. Moving it up and changing its colour would make it far more visible."

Flow
  1. Email template
  2. Reading view
  3. Attention analysis
  4. Email score
  5. Suggested edits
CTA attentionDistractionCognitive load
S-04

Creative Intelligence Reports

What do the best ads in your category look like?

The problem
Brands lack industry attention benchmarks to know where they stand.
How it solves it
Periodic reports with attention benchmarks by category, trends (video vs. image, rising colours) and a ranking of the best neurologically performing assets.
What your team gains
Market intelligence in three tiers: free summary, full version and custom category edition.
Flow
  1. Most-run ads
  2. Category analysis
  3. Benchmarks and trends
  4. Tiered report
BenchmarksTrendsRanking
S-05

Packaging War Room

Is your product visible on a shelf with 40 competitors?

The problem
Packaging that shines in isolation can vanish on a three-metre shelf full of competitors.
How it solves it
From a photo of the real shelf, it detects each product, overlays an attention heatmap and quantifies what share of attention your product captures against each competitor. Then it proposes changes and projects their impact.
What your team gains
Packaging and display decisions based on real attention at the point of sale.
Flow
  1. Shelf photo
  2. Product detection
  3. Attention map
  4. Attention share
  5. Before / after mock-ups
Attention shareCompetitorsDelta
S-06

Creative Due Diligence

Is this brand growing on brand strength or on discounts?

The problem
Those evaluating an investment in a brand need to know whether its communication builds the brand or just pushes price.
How it solves it
It assesses the brand’s creative against category benchmarks: average attention, brand visibility and relative position.
What your team gains
An independent diagnosis, with a written report and a presentation with the expert panel.

Illustrative example"Focus at the 22nd percentile of its category and low brand visibility: growth comes from price, not brand strength."

Flow
  1. Brand assets
  2. Category benchmarks
  3. Percentile and visibility
  4. Diagnosis
PercentileBrand visibility
WHO IT'S FOR

Teams that invest in media and want to know why it works.

Let's bring your next campaign to the lab.

We run an assisted pilot with your creatives and your media data.

Let's talk