Does it survive the first three seconds?
“Will the viewer’s thumb stop, or scroll past before the brand even lands?”
Pre-spend creative intelligence
Score the ad,
before you spend.
Upload one image or video. Get a verdict, five KPIs, an attention heatmap and a synthetic buyer panel in ninety seconds. Calibrated against 1,200+ public outcome ads. Built by operators from NielsenIQ and L'Oréal.
★★★★★Used by performance marketers at DTC brands and agencies·First three ads free
app.saliencylab.com · ScoringLive
Beat the Skip82
Sell Proposition61
Brand Impact88
Verdict
Scale
Composite
76/100
Built by experts
Operators who lived the creative testing problem from the inside.
Operators already running pre-spend diagnostics with SaliencyLab
Brand directors, paid-media leads, and founders across L'Oréal, Estée Lauder, and YC-backed startups.
Running L’Oréal Paris in Morocco means balancing global brand codes with local cultural nuance. SaliencyLab’s MENA-aware benchmark finally gives brand managers here the same creative intelligence that global teams in Paris have had for years.
Brand Director, L’Oréal Paris Morocco
Casablanca, Morocco
I’ve bought paid media from Beirut to Rome, across Amazon, Instagram, Pinterest, and LinkedIn, and every channel has its own creative failure mode. SaliencyLab is the first diagnostic I’ve used that flags what’s wrong before the data comes in, when I can still change the cut.
Senior Media Manager, FBU Solutions
Rome, Italy
I’ve built products on YC’s playbook: validate cheap, ship fast. SaliencyLab is how I validate creative without briefs or wait times. And as a Moroccan founder, watching the MENA benchmark deepen every week genuinely matters.
Co-founder & CTO, Mailwarm (YC S20)
Dubai, UAE
← swipe to read more →
How it works
Score, decide, ship, in 90 seconds.
Four steps. Zero surveys. The same framework whether you're testing one ad or a hundred variants.
01
↑
Drop your ad here
MP4 · MOV · JPG · PNG · ≤ 60s
📹 demo-video.mp4 · 8.4 MB
01
Upload one ad. Any format.
Video or image. MP4, MOV, JPG, PNG, up to 60 seconds. No SDK, no integration, no media-buying-platform setup. Just drag, drop, done.
≤ 5 sec to upload02
RoastIQ ANALYZING
0:00 · Coca-Cola bottle · 99%
02
RoastIQ scores it scene by scene.
RoastIQ reads the visuals, transcribes the audio, classifies 43 creative attributes per scene, and benchmarks the result against 1,200+ real ads. Every detection is timestamped and traceable.
90 seconds end-to-end03
Coca-Cola · Holidays 2025
SHARPEN81Beat
Skip
Skip
79Get
Noticed
Noticed
82Brand
Impact
Impact
63Sell
Prop
Prop
81Build
Brand
Brand
77
Composite→ Next move
Tighten the value proposition before second 8. The emotional hook earns attention but the product reason arrives too late.
Tighten the value proposition before second 8. Hook earns attention; product reason arrives too late.
BENCHMARK: 142 ADS · YT · DTC SKINCAREP75, TOP QUARTILE
03
5 KPI scores + a single verdict.
Scale, Sharpen, or Rebuild. Each verdict is tied to specific creative attributes (the hook timing, the brand's visual prominence, the sell proposition pacing) so you know what to change, not just whether to ship.
Held-out OOS ρ +0.3204
04
When the score still leaves the ‘why’ unanswered, open BuyerLens.
Synthetic users built from your category, market, and target persona explain their resistance, in their own voice. Use the answers to shape the next edit, not just guess at it.
Optional · Opens from any RoastIQ resultTwo tools. One decision.
Score first. Then ask why it lands, or does not.
RoastIQ delivers the verdict in ninety seconds. BuyerLens opens from that verdict to simulate the buyer's voice. The friction, the doubt, the missed cue. One project layer connects them.
01
RoastIQ · The verdict
Score any ad in ninety seconds.
Pre-spend diagnostic. One image or video in, verdict out.
Five frozen KPIs. A held-out benchmark of public outcome ads. A predicted attention heatmap. The evidence trace that names every win and every fix.
- Verdict: Scale, Sharpen, or Rebuild
- Five KPIs across skip, notice, brand, sell, build
- Calibrated against 1,200+ public outcome ads
02
BuyerLens · The resistance
Simulate who pushes back, and why.
Synthetic personas, opens from a RoastIQ result, never standalone.
A 36-persona synthetic panel reacts to your creative in their own voice. Surface objections, missed cues, brand fit friction, before the next edit, not after the next €200k.
- 36 grounded personas across markets and segments
- Tagged objections: price, fit, claim, novelty, brand
- One next move, the edit that flips the most resistance
Methodology
One composite. Five answerable questions.
Every verdict decomposes into a single 0 to 100 composite and five scored questions you can act on. No black box brand equity index. Every weight is held out cross validated against public outcome data.
Composite0–100CV tuned · platform native
Scale · ≥ 70Sharpen · 55 to 69Rebuild · < 55
BuyerLens · The simulation
Hear the resistance before you ship the next cut.
From a RoastIQ verdict, BuyerLens runs a synthetic panel of 36 grounded personas against your creative. You hear the doubt, the misread, the brand fit friction in the buyer's own voice.
It is not a chat with an LLM. It is a panel that talks back.
Every persona is grounded in market, segment, and category. Every response is tagged for the objection class: price, fit, claim, novelty, brand. The output is one next move, the edit that flips the most resistance.
i
Inherit context from RoastIQ
No fresh setup. Category, goal, target segment carry across.
ii
36 personas react in their own voice
Tagged objections, structured reasoning, never generic.
iii
One next move
The single edit that lifts the most predicted resistance.
app.saliencylab.com/buyerlens · Study in progressLive
A
Price doubt“It's beautiful, but the closing frame doesn't tell me why this costs more than the one in my bathroom.”
J
Claim friction“The promise is fast, but the demo skips the texture. I'd scroll past at the 4 second mark.”
K
Brand fitformulating response…
Panel · 22/36 personas62%
Validation
Calibrated against real ads, not vibes.
Every score is benchmarked against a held-out cohort of public outcome ads. We do not validate against customer-reported ROAS. Too noisy, too unfalsifiable. We validate against TikTok engagement, TikTok CTR, and YouTube view counts.
1,200+
Ads in the held-out calibration cohort
ρ +0.32
YouTube view counts · OOS Spearman, n=403
ρ +0.31
TikTok engagement · OOS Spearman, n=700
6.5×
Top-quintile vs bottom-quintile lift
Why we built this
Pre-spend creative testing has been broken for thirty years.
The $56B global market-research industry still runs on consumer surveys: small samples, slow turnaround, no methodology disclosure. Kantar alone has tested 230,000+ ads through its Link survey panel. SaliencyLab is the alternative, multimodal AI scored against actual public outcomes.
Sources: ESOMAR Global Market Research 2024 (insights industry $150B, market research sector $56B). Kantar Link 230,000+ ad database disclosed in Kantar Creative Effectiveness 2024.
The old way
Survey-based testing
- 4–8 weeks from brief to scorecard
- Survey panel of ~150 respondents per cell, small N, biased toward survey-takers
- No methodology disclosure, proprietary "black box" scoring
- Concurrent validity reported, never held-out validation
- Built for the era of TV. Optimized for one ad at a time.
- Anchored to recall surveys, what people say they remember, not what the feed actually rewards.
The SaliencyLab way
Pre-spend AI scoring
- ~90 seconds from upload to scored verdict
- Calibrated against 2,000 real ads with public engagement & click-intent outcomes
- Open methodology, code public, peer-review track at Scientific Data
- Held-out 5-fold CV: ρ +0.30 to +0.32 across TikTok engagement, CTR, YouTube views
- Built for the era of feed. Optimized for fast iteration across many variants.
- Anchored to 42M behavioral signals, what the feed actually rewards, not survey self-report.
Does the score predict sales or ROAS?
Truth-first
No. RoastIQ predicts relative engagement and click-intent on TikTok and YouTube. It is calibrated against the public outcomes of 1,200+ real ads (held-out OOS Spearman ρ +0.30 to +0.32).
We do not claim sales lift, ROAS, attributed conversion, or brand recall. The score is a pre-spend gate, not a substitute for in-market measurement. That honesty is part of the contract.
How is this different from Kantar Link AI or other survey-based testing?
No survey panel. No three-week wait. No proprietary black box.
SaliencyLab is calibrated against 1,200+ public-outcome ads with measured engagement and click-intent data. Methodology is open, validation code is reproducible, and every report shows the benchmark cohort size for your category. Ninety seconds instead of three weeks, at a fraction of the cost.
How accurate is it, really?
Held-out 5-fold cross-validation across three signals: TikTok engagement (ρ +0.31, n=700), TikTok CTR (ρ +0.30, n=691), YouTube view counts (ρ +0.32, n=403).
Pool-wide quintile lift is 6.5×. Ads in the top predicted quintile beat the bottom quintile by that multiple on real engagement. Scaling curves show ρ still growing as the cohort expands. Meta is directional defaults only; we mark it as such on every Meta-Feed score.
Is my creative private?
Your creative stays in your workspace. We do not use customer ads to train scoring models. Calibration runs on the public benchmark cohort only.
Storage is purged on request. GDPR-ready. SOC 2 Type II in progress. DPA available on the Enterprise tier.
Does it work for static images, or only video?
Both. MP4, MOV, GIF, PNG, JPG up to 60s.
Static images skip the audio and video-saliency stages and finish in about thirty seconds. Long-form video beyond 60s is on the roadmap.
What if my category isn't in the benchmark yet?
Every report shows the benchmark cohort size next to the verdict. If your category has fewer than 100 ads in the pool, we flag it as “Building” with the current count and growth rate, never hidden as “Unbenchmarked.”
If you need a private brand benchmark, the Pro tier builds one across your own uploads so you can track drift over time even where the public pool is still growing.
Pricing
Start free. Pay when you scale.
First analysis in under ninety seconds. No card required. Same scoring engine across every tier. We do not gate accuracy by plan.
Free
€0/ month
For curiosity, brief reviews, and early stage creative checks.
- 3 analyses per month
- Full 5 KPI breakdown plus verdict
- Platform native scoring
- Public benchmark comparison
- Methodology disclosure on every report
Most chosen
Pro
€99/ month
For brand teams iterating on creative every week. Unlimited analyses plus BuyerLens.
- Unlimited analyses
- BuyerLens synthetic buyer panel
- Brand drift tracking over time
- Private benchmark cohort
- Priority compute · sub 60 second scoring
- Team workspace · 5 seats included
Enterprise
Custom
For agencies and global brands needing private benchmarks, API access, SOC 2.
- Everything in Pro
- Private benchmark cohorts
- API access plus webhooks
- Dedicated CSM and onboarding
- SOC 2 Type II plus DPA
- Volume pricing for high throughput teams
14 day free trial on Pro · No card required · Cancel anytime · GDPR ready · SOC 2 Type II in progress
Book a demo · 30 min
Stop guessing.
Start simulating.
Thirty minutes. We score one of your real ads live and walk through the verdict, heatmap and synthetic buyer reaction. You leave with the report, even if you do not sign up.
★★★★★No slides, real product·Founder-led demo
Oussama Nakhil
Founder at SaliencyLab. Previously L'Oréal Groupe, NielsenIQ, HEC Paris, Grenoble EM.
SaliencyLab · Demo
SOC 2 Infrastructure
GDPR Compliant
256-bit Encryption
Row-Level Isolation
No Cross-Customer Data Sharing