Ad creative testing platforms are tools that measure how a real or simulated audience responds to advertising creative — before or after it runs — so teams can optimize for attention, emotion, recall, and purchase intent before committing full media spend. The best platforms go beyond simple surveys to capture behavioral and emotional signals that survey data alone cannot reveal.

How we evaluated these platforms

• We reviewed platforms whose core job is testing or scoring ad creative — not general survey tools or media planning software. Platforms were assessed across six criteria:

• Measurement depth — does it capture one signal type or multiple?

• Signal types captured — does it cover what consumers say, what they do (eye tracking, attention), and what they feel (emotional response)?

• Speed to insight — hours vs. days vs. weeks

• Format support — video, static image, packaging, digital, OOH

• Pricing transparency — is pricing publicly listed or enterprise-only?

• Best-fit team — startup growth team, enterprise CMO, creative agency, or performance marketer?

Transparency note: Decode by Entropik is Entropik's own platform and holds Position 1 in this listicle. It was evaluated against the same criteria as every other platform. We placed it first because it is the only pre-launch tool in this list that measures all three signal layers (say, do, and feel) simultaneously — a factual distinction, not promotional placement.

The two types of ad creative testing platforms

Before comparing specific tools, it helps to understand the structural divide in this category. Most teams that struggle with creative testing are actually mixing up two completely different types of tools — and using them at the wrong stage.

Type 1 — Pre-launch pretesting

Pre-launch pretesting measures how a real or modeled audience responds to your creative before you commit media budget. The goal is to eliminate weak concepts, identify emotional engagement gaps, and rank creative options when you can still act on the findings.

Representative platforms: Decode by Entropik, Kantar LINK+, System1 Test Your Ad, Dragonfly AI, Neurons.

Type 2 — In-market testing

In-market testing runs live budget experiments after launch. You serve different creative variations to real audiences with real spend and measure which version drives better results on actual KPIs (CTR, conversion, ROAS).

Representative platforms: Marpipe, Motion, Meta A/B testing, Smartly.

Why mature teams run both

Pre-launch testing catches disasters before they scale. In-market testing validates real-world performance and catches diminishing returns before you burn out an asset. The two approaches answer fundamentally different questions:

Pre-launch: "Which of these 3 concepts should we produce?"

In-market: "Which version of the produced ad should we scale?"

Using only one is like checking the weather forecast but never packing a coat.

Platform taxonomy at a glance

A note on creative's contribution to advertising effectiveness: A meta-study of approximately 450 CPG campaigns by NCSolutions found that creative quality drives roughly 49% of advertising's incremental sales impact — more than targeting, reach, or recency combined. A separate finding from Kantar's Link database shows the most creatively effective ads generate approximately 4x as much profit as average executions. Getting creative right before launch is one of the highest-ROI optimizations available to most teams.

10 Best ad creative testing platforms compared

The table below maps every major platform in this guide against the three signal layers: SAY (verbal/survey), DO (behavioral/eye tracking/attention), and FEEL (emotional/facial coding).

The say/do/feel column split is the most useful filter: if you need to know why an ad performs, not just whether it performs, you need pre-launch tools with behavioral and emotional signal coverage.

1. AI Creative Insights by Decode

Testing stage: Pre-launch

Formats: Video, static image, packaging, digital display, OOH

Best for: CPG brands, FMCG companies, BFSI, enterprise insights teams running multi-concept creative tests

Decode's AI Creative Insights product is the only platform on this list that simultaneously measures what respondents say, what they do (behavioral attention), and what they feel (emotional response) — the complete three-signal picture required to understand creative performance before a single dollar of media runs.

What Decode measures

• SAY (verbal layer): Post-exposure survey responses, recall probes, brand linkage, messaging comprehension. Standard across the category.

• DO (behavioral layer): Eye tracking captures exactly where attention lands on each frame. Heat maps show which design elements draw focus, which are ignored, and whether the brand or product registers in the first three seconds — the window that determines whether an ad earns continued attention.

• FEEL (emotional layer): Facial coding analyzes micro-expressions across 62 facial action units during exposure, mapping an emotion curve across every second of the ad. Voice AI captures sentiment in verbal responses beyond what words alone convey. This is the layer no survey can replicate.

Verified platform figures:

• 90%+ facial coding accuracy

• 96% eye tracking accuracy

• 62 facial expressions tracked

• 70+ languages supported

• 17 patents in emotion AI

• 150+ global brands served

• SOC2 Type II + ISO 27001 + GDPR compliant

• 100M+ panel reach (Cint/Dynata)

A real-world scenario

A CPG team is testing three 30-second TV ad concepts before green-lighting one for full production. With traditional pretesting, this takes 4–6 weeks. With Decode's AI Creative Insights:

1. Concepts are uploaded and distributed to matched audience panels

2. Participants watch on desktop (facial signals are more reliable on desktop camera)

3. Within hours, the team receives frame-level attention heatmaps, an emotion curve over time, brand recall scores, and verbatim feedback

4. They can see exactly where Concept A loses emotional engagement (second 12), where Concept B recovers attention after a dip, and which concept drives the strongest brand linkage

The result: a data-driven production decision made in days, not weeks — using all three signal layers.

Honest limitation: Decode's AI Creative Insights is a pre-launch behavioral research platform. It is not an in-market spend-optimization or creative generation tool. Teams looking to iterate on live creative variations with real budget should pair it with an in-market platform like Marpipe after launch.

Test your creative with Decode →

2. Kantar LINK+

Testing stage: Pre-launch

Best for: Large enterprise brand teams that need benchmarks against an established global database

Kantar LINK+ is one of the most established pre-launch creative pretesting systems in the industry. Its primary differentiator is database depth: decades of normed results across hundreds of thousands of campaigns, categories, and markets that allow brands to compare a new asset against what has historically worked in their category.

Strengths:

• Extensive global benchmark database across CPG, automotive, finance, and FMCG — among the largest in the industry

• Automated results delivery for faster turnaround on standardized ad tests; largely survey-based, samples approximately 150 respondents per test

• Widely trusted by enterprise CMOs and media planners for justifying production investment

• Strong coverage of TV + video; growing digital capabilities

• Results formatted for executive-level reporting and cross-market comparisons

Honest limits:

• Largely survey-based with limited behavioral nuance; real-time biometric signal capture (facial coding, eye tracking) is not the primary methodology

• Sample sizes of approximately 150 reflect standardized norm-comparison models, not statistically powered individual brand studies

• Premium pricing structure that puts it out of reach for most challenger brands or mid-market teams

• Slower turnaround than AI-native behavioral platforms when custom methodology is needed

Kantar LINK+ is the right choice when you need historical market norms and executive credibility behind your creative decision. It is not the right choice when you need same-day behavioral feedback on three rough-cut variants.

3. System1

Testing stage: Pre-launch

Best for: Brand teams focused on TV and digital advertising who want emotion-based market norms and long-term brand equity prediction

System1 Test Your Ad is built around the principle that emotional response — specifically the intensity and type of feeling an ad generates — is the strongest predictor of long-term brand equity growth. Its 1–5 star effectiveness rating (the "Star Rating") maps to predicted market share growth and provides an accessible, defensible metric for creative decisions.

Strengths:

• Emotion-based scoring methodology grounded in behavioral science research

• Designed to distinguish between short-term sales activation and long-term brand equity — a useful distinction for brand-building briefs

• Market norms database for TV and digital advertising; useful for cross-category benchmarking

• Accessible interface for non-research teams; dashboard-driven reporting

Honest limits:

• Emotion detection approach uses survey-assisted methods; distinct from the biometric AI used by platforms like Decode (facial coding, eye tracking with real participants)

• Less granular second-by-second behavioral tracking compared to direct biometric measurement

• Primarily optimized for TV and video; static and packaging support is more limited

• No real-audience behavioral signal layer (eye tracking, attention measurement)

System1 Test Your Ad is strong for teams prioritizing emotional effectiveness benchmarks and brand equity prediction. It works best alongside behavioral platforms when your brief also requires attention data or packaging format coverage.

4. Dragonfly AI

Testing stage: Pre-launch (predictive, no real participants)

Best for: In-house creative teams and agencies needing instant attention feedback on static and video creative without running a study

Dragonfly AI uses computational models trained on eye-tracking datasets to predict where human attention will fall on an image or video frame. Upload a creative asset and within seconds receive an AI-generated attention heatmap — without recruiting a single participant.

Strengths:

• Instant feedback — seconds, not hours or days

• No participant recruitment required; self-service and scalable for high-volume creative teams

• Useful for rapid creative iteration during the design phase before briefing a full study

• Good at catching obvious attention failures: logo placement, headline legibility, visual hierarchy

• Accessible pricing relative to full behavioral study platforms

Honest limits:

• Predictive models, not real participant data — attention is modeled from training datasets, not measured from actual audience responses to your specific creative

• Attention only — no emotion signals, no survey data, no real-world audience reactions

• Cannot tell you how an ad makes someone feel, only where their gaze might land

• Predictive accuracy may vary for novel creative executions or culturally specific content not well-represented in training data

Dragonfly AI is best used as a fast filter during creative development — a way to identify and fix obvious visual problems before investing in a full study. It is not a substitute for real behavioral research with live audiences when the brief requires emotional or persuasion data.

Checkout: Decode vs Drangonfly AI

5. Neurons

Testing stage: Pre-launch (predictive, no real participants)

Best for: Agencies and design teams that need attention heatmaps plus cognitive engagement estimates for digital advertising and visual media

Neurons applies neuroscience-informed predictive modeling to generate attention heatmaps, cognitive load estimates, and engagement scores — without live participant panels. Its model is trained on eye-tracking and neuroscience datasets and produces predictions fast enough to fit into active design iterations.

Strengths:

• Attention heatmaps plus cognitive load and memory encoding estimates — a slightly broader signal set than pure attention tools

• Fast enough to run multiple creative variants side-by-side within a single design session

• Covers digital ads, packaging, and visual hierarchy in web pages or app interfaces

• Used by agencies and brand teams for digital ad refinement at speed

Honest limits:

• Model-predicted rather than directly measured per audience — the same foundational limitation as Dragonfly AI

• No emotional signal layer; no survey data; no real participant responses to your specific creative

• Engagement and memory scoring are model-estimated, not measured from actual brain activity

• Cultural nuance and audience-specific attention patterns may not be fully captured by general training data

Neurons is best for design and creative teams that need fast, structured attention feedback across multiple iterations. Like Dragonfly AI, it is a development aid — not a substitute for real-audience pretesting on high-investment campaigns.

Quick Read: Decode vs Neurons

6. Marpipe

Testing stage: In-market

Best for: Performance marketing teams running paid social who want to identify winning creative combinations with real spend

Marpipe's core workflow is clear: build a matrix of creative variations (headline A vs. B, image A vs. B, CTA A vs. B), run them as real paid ads against real audiences, and use spend-weighted data to identify which combination wins. Everything is measured against actual in-platform performance metrics: CTR, conversion rate, ROAS.

Strengths:

• Tests real creative combinations with real audience responses and real business outcomes

• Isolates which creative variables drive performance — tells you why a variant won, not just which won

• Clear winner/loser metrics tied directly to ad platform performance

• Good interface for managing large test matrices; connects directly to ad platforms (Meta, etc.)

Honest limits:

• In-market only — requires live budget; cannot replace pre-launch testing to avoid producing weak creative in the first place

• No pre-launch behavioral data — no way to understand why a variant might outperform before launch

• No emotional signals — no facial coding, eye tracking, or survey-based sentiment

• Findings are platform-specific; a winner on Meta does not necessarily translate to YouTube or CTV

Marpipe answers the post-production optimization question. It does not replace the pre-production decision.

7. Motion

Testing stage: Post-launch analytics

Best for: DTC brands, performance marketing teams, and social advertising teams managing large creative libraries

Motion connects to your ad accounts and surfaces creative performance data in a way that makes it easy to identify top performers, spot creative fatigue, and understand which creative elements correlate with results over time. It is primarily an analytics layer — it does not generate creative or launch campaigns.

Strengths:

• Strong visualization of post-launch creative performance data across a large creative library

• Identifies creative fatigue before it destroys performance; surfaces patterns over time

• Helps performance teams attribute outcomes to specific creative elements

• Integrates with Meta, TikTok, YouTube, and other platforms

• Popular with DTC and e-commerce brands managing dozens of concurrent creatives

Honest limits:

• Analytics, not testing — Motion tells you what happened to creative that already ran; it cannot predict performance before launch

• No emotional signals, no behavioral signals, no pre-launch research capability

• Most useful when you already have significant creative performance history; less valuable for new brands or new categories with thin data

Motion belongs in the post-launch layer of a mature creative testing workflow. It enhances a methodology; it does not replace one.

8. AdCreative.ai

Testing stage: Pre-launch (generation + predictive scoring)

Best for: High-volume ad teams, SMB marketers, and agencies that need rapid AI-generated creative with basic performance scoring

AdCreative.ai uses generative AI to produce ad creative (static images, banners, copy variations) and pairs it with a proprietary "Creative Score" that predicts performance against benchmarks. For teams running high-volume digital campaigns with limited production resources, it collapses the production and initial screening steps into one workflow.

Strengths:

• Fast AI creative generation at scale; useful for performance campaigns where volume matters

• Integrated scoring gives teams a starting point for creative decisions before spend

• Accessible pricing relative to full research platforms; scales well for SMB and agency use cases

• Covers static ads, social formats, and copy generation

Honest limits:

• A generation tool first; the scoring is a model-based estimate, not a real-audience behavioral study

• No live audience research — no eye tracking, no facial coding, no survey responses from real participants

• Scoring methodology is not as transparently documented as established research platforms

• Best suited as an efficiency tool for commodity ad production, not as a substitute for rigorous pre-launch behavioral testing on high-investment brand campaigns

For teams balancing production efficiency with basic pre-launch screening, AdCreative.ai helps on the production side. It does not replace the testing side for brand-critical campaigns.

9. Pencil

Testing stage: Pre-launch (generation + predictive scoring)

Best for: Performance marketing teams and enterprise brands running high-volume video advertising who need AI-assisted creative production at scale

Pencil uses generative AI to produce video ad variants from existing creative assets — scripts, visuals, voiceover — and predicts which versions are likely to perform based on historical in-platform performance patterns. It is designed for teams that need to generate and score many video variations quickly, particularly for social platforms like Meta, TikTok, and YouTube.

Strengths:

• AI-assisted video variation generation at scale; creates multiple versions from a single asset set

• Predictive performance scoring before launch, drawing on historical ad performance benchmarks

• Designed for video-heavy campaign workflows; useful for DTC and performance teams with high creative volume

• Reduces production bottlenecks when testing many variations is required

Honest limits:

• A creative production tool with scoring functionality — not a behavioral research platform

• Predictive scores are based on historical performance patterns, not live audience emotional or attention measurement

• No facial coding, eye tracking, or real participant survey methodology

• For brand-critical or emotionally complex creative, performance scoring models are not a substitute for real behavioral pretesting

Pencil is best suited for performance teams managing high volumes of video creative who need production efficiency alongside basic pre-spend guidance. Teams investing in major brand campaigns should use a behavioral pretesting platform in parallel.

10. Omneky and CreativeX

Omneky is an AI-powered creative generation and optimization platform for digital advertising. It generates branded ad variants at scale and connects performance feedback to future creative decisions — creating a feedback loop between live campaign data and new creative production. Best for enterprise performance teams managing large cross-channel creative libraries across multiple markets.

CreativeX is a creative intelligence platform focused on measuring creative quality and consistency at scale. Rather than testing with audiences, it analyzes existing creative assets against brand guidelines, effectiveness markers, and cross-market consistency standards. Best for global brand governance teams responsible for maintaining creative quality and compliance across a large, distributed organization.

Honest limits for both: Neither platform captures real audience behavioral or emotional response (no facial coding, eye tracking, or live participant research). They are data infrastructure and governance tools, not audience research platforms. Teams that need to understand why creative works — not just whether it meets brand standards or generates variants efficiently — should pair these tools with a pre-launch behavioral platform.

How to choose an ad creative testing platform

Match the platform to the question you actually need to answer, the stage you are at in the creative process, and the depth of signal your brief requires. Use this decision guide:

If your priority is pre-launch behavioral testing with real audiences (SAY + DO + FEEL):

• Decode by Entropik — the only platform in this list that measures all three signal layers simultaneously with real participant panels. For enterprise norm benchmarking alongside

If your priority is instant attention feedback during creative development (no participants):

• Dragonfly AI or Neurons — predictive attention models; no participant recruitment needed; results in seconds. Note: attention-only signal, no emotional data.

If your priority is finding the winning creative variation with real in-market spend:

• Marpipe — multivariate in-market testing with real budget and real performance signals. Best deployed after pre-launch testing has eliminated weak concepts.

If your priority is monitoring creative performance and fatigue across a live campaign library:

• Motion — post-launch analytics connecting creative attributes to performance outcomes over time.

If your priority is high-volume AI video creative generation with basic scoring:

• Pencil — AI video variation generation with predictive performance scoring. For static and banner formats: AdCreative.ai.

If your priority is enterprise creative governance and quality consistency across markets:

• CreativeX — cross-market brand compliance and creative quality analysis. For generation at scale: Omneky.

A common sequencing mistake: Teams often choose an in-market testing tool (Marpipe, Motion) and skip pre-launch behavioral testing — then wonder why their "winning" variation still underperforms against benchmarks. Pre-launch behavioral testing answers should we produce this? In-market testing answers which version should we scale? Both questions matter; they need different tools.

The research-to-spend gap: Research suggests marketers significantly underestimate how much creative quality drives advertising outcomes. Pre-launch testing is often where the highest ROI gains are available, because the cost of changing creative before production is low; the cost of scaling the wrong creative is high.

For the methodology behind how to structure a pre-launch creative test — including what stimuli to use, how to recruit the right audience, and how to interpret emotional curves.

How do AI creative testing platforms compare to traditional focus groups?

AI creative testing platforms and traditional focus groups are not direct substitutes — they answer different questions with different trade-offs. Understanding where each method wins is more useful than treating them as alternatives.

What focus groups still do better: When you need to understand the story behind a reaction — why a character choice feels off-brand, what cultural association the tagline activates, what the creative makes people think about your category — skilled qualitative moderation produces narrative insight that automated systems cannot match.

• What AI creative testing platforms do better: When you need statistically robust emotional and attention data across a representative audience, on a timeline that fits a production schedule, AI creative testing platforms are faster, larger in sample, and capture signals (involuntary micro-expressions, second-by-second emotion curves, gaze patterns) that focus group discussion cannot surface.

  
  

• The emerging middle path: AI-moderated qualitative research platforms — such as Mira, Decode's AI Moderated Interview product — run AI-moderated discussions with behavioral signal measurement simultaneously. This combines the depth of a moderated conversation with the scale and signal richness of automated behavioral coding. For creative testing briefs that require both depth and quantification, this approach is worth evaluating. For a broader overview of AI moderated research methods, see the NN/g overview of AI interviewers

  
  

• The honest summary: AI creative testing supplements traditional qualitative research for most briefs — it does not always replace it. Use behavioral platforms to validate and scale at speed; use moderated qualitative to go deeper when findings surface unexpected directions.

Can AI predict ad performance without testing on real people?

This question sits at the center of a real tension in the category: predictive AI tools (Dragonfly AI, Neurons, AdCreative.ai, Pencil) promise fast performance signals without participant recruitment. Real-audience platforms (Decode, Kantar LINK+, System1) require actual respondents. Both approaches are called "AI creative testing" — but they are fundamentally different methods with different confidence levels.

Where predictive models work well:

Attention prediction models (Dragonfly AI, Neurons) perform reliably for what they are designed to do: identifying where the eye is likely to fall on a well-structured image or video frame, based on saliency patterns consistent across most audiences. For visual hierarchy checks — "Is the product visible in the first three seconds?", "Does the headline compete with the background?" — predictive models are fast, accessible, and sufficient for development-stage feedback.

Predictive scoring in generation tools (AdCreative.ai, Pencil) can filter out obvious underperformers before a single impression is served, based on patterns in historical ad performance databases.

Where predictive models fall short:

Attention prediction models do not measure emotion. They cannot tell you whether your ad creates anxiety, delight, trust, or confusion. They model where people might look, not how people feel. For ads where emotional resonance is the primary effectiveness driver — brand-building creative, product launches, emotionally complex campaigns — predictive attention alone is not a sufficient signal.

Scoring models trained on historical performance data may also miss best-performing creative in new categories, novel executions, or culturally specific markets where training data is thin.

The Decode approach:

Decode's AI Creative Insights uses real audience panels — real people watching the actual ad, with eye tracking and facial coding capturing their real responses in real time. The AI processes and analyzes those responses at scale; it does not replace the real-human signal with a predicted one. This matters for high-investment brand campaigns where you need to know how your actual target audience actually responds, not how a model predicts a generalized audience might respond.

The practical answer: Predictive AI models are a fast, cost-effective filter during creative development. They are not a substitute for real-audience behavioral research when the stakes are high. Use predictive tools early and often during development; use real-audience platforms before committing significant production or media investment.

How Decode helps

Most platforms pick a lane: they measure what people say, or they model where attention might fall, or they track post-launch performance metrics. Decode is built for teams that need the complete picture — all three signal layers, with real audiences, before the media budget is committed.

The AI Creative Insights product sits within Decode's broader unified platform, which also covers AI Moderated Interviews (Mira), Consumer Insights, User Research, and Insights Hub — so creative test findings connect directly to other brand and product research without switching platforms or reconciling disconnected data.

The recommended workflow: pretest with Decode's behavioral platform to eliminate weak concepts and optimize before launch, then let in-market tools (Marpipe, Motion) optimize among the produced assets that go live.

For teams evaluating creative testing platforms in 2026, the key question is no longer whether to test — it's whether your testing methodology captures enough signal to make confident decisions. Survey scores alone cannot tell you that your ad lost emotional engagement in the second half, that your logo isn't registering in the first three seconds, or that your target audience feels anxious rather than excited during your product reveal. Those findings require behavioral and emotional measurement with real participants.

Frequently asked questions

1. What is the best ad creative testing platform?

The best ad creative testing platform depends on what stage of the creative process you're in and what signals matter most to your team. For pre-launch behavioral testing that covers emotional response, attention, and verbal feedback in a single study, Decode by Entropik is the strongest option. For in-market multivariate testing with real spend, Marpipe is the category leader. For predictive attention checks without participant recruitment, Dragonfly AI and Neurons are the fastest options.

2. How does AI creative testing work?

AI creative testing uses artificial intelligence to automate parts of the research process — including participant matching, facial expression coding, attention modeling, and insight synthesis. In behavioral platforms like Decode, AI enables real-time facial coding (analyzing micro-expressions during ad exposure) and eye tracking (measuring where attention falls). In predictive tools like Dragonfly AI, AI generates attention heatmaps from computational models without needing live participants. The depth of insight varies significantly between the two approaches: real-audience behavioral AI produces richer, more reliable data than predictive models alone.

3. What is the difference between pre-launch and in-market creative testing?

Pre-launch creative testing measures audience response to ad creative before it goes live — using real or modeled audiences in a controlled environment. It answers questions like: "Which of these concepts is most emotionally engaging?" and "Will our target audience notice our brand in the first three seconds?" In-market testing runs live creative variations with real paid media spend and measures actual downstream performance — CTR, conversion rate, ROAS. Pre-launch testing prevents weak creative from going to market; in-market testing optimizes among produced creative to find the best performer. Both have a role in a mature creative testing workflow.

4. How much does creative testing software cost?

Pricing varies significantly by platform type and tier. Predictive attention tools (Dragonfly AI, Neurons) typically offer subscription plans starting from a few hundred dollars per month. Automated ad testing platforms (AdCreative.ai) range from self-service tiers at a few hundred dollars per month to enterprise plans at several thousand. Full behavioral research platforms (Decode by Entropik, Kantar LINK+) are enterprise-priced based on study volume, market, and contract terms — contact vendors directly for current pricing. In-market tools (Marpipe, Motion) are generally SaaS subscriptions with pricing based on ad spend under management or feature tier. System1 Test Your Ad Pro is priced per ad test at enterprise rates.

5. What signals should I measure in creative testing?

The most comprehensive pre-launch creative testing captures three signal layers: SAY (what respondents tell you in surveys — recall, brand linkage, message takeout), DO (what they actually pay attention to — eye tracking, attention heatmaps, time-to-first-fixation), and FEEL (how they emotionally respond — facial coding, emotion curves, voice sentiment). Most platforms cover SAY. A subset covers DO. Very few cover all three. The more signal layers your test captures, the more confidently you can diagnose why a piece of creative will or won't work — not just whether respondents say they like it.

Related Topic:

• Best Consumer Insights Tools for Market Research in 2024

• Best Consumer Research Platforms in 2026

• Best Usability Testing Tools for UX Research Teams

• Top 10 Customer Journey Mapping Tools for Market Research

• Top 10 Survey Tools in 2024: A Comparison for Research Teams