A platform to capture mental models from thousands of stakeholders. Create surveys, collect responses, and discover patterns—all in one place.
Companies spend billions on surveys, focus groups, and consultants—but they only get opinions. They don't understand how people are organizing information in their heads.
Example: A survey tells you "60% support climate policy." But why? Some see economic benefits. Others see moral duty. Others see system collapse. Same opinion, completely different thinking.
Surveys ask questions you think are important. They don't show you what stakeholders think is important and how it all connects. You can't design effective messages without understanding their mental structure.
Get Causality measures how people organize information—their mental models at scale
We're a small independent team working hard to meet growing demand. Tell us about your use case so we can prioritize onboarding based on your needs.
Join the waitlist for beta access and founding member pricing
Try the live demo and see how it works
We measure how people organize information in their heads—their mental models. Surveys tell you opinions. We show you how they're thinking about the problem.
A mental model is how someone internalizes cause-and-effect relationships. It's their personal theory of how things work—and different people can have completely different mental models about the same topic.
Sign up (free tier available) → Create a survey or import data → Share a link → Platform aggregates responses → Run analyses with one click → Export publication-ready results.
Yes! The free plan gives full access to core features with no credit card required. You can also try the live demo or join the beta waitlist for early access.
The only platform offering end-to-end research: from concept design to large-scale data collection to AI-powered analysis
Facilitated concept elicitation with stakeholders. Design your FCM survey instrument.
Explore Workshops →Deploy FCM surveys to thousands. Collect mental models at scale with our specialized survey builder.
See Survey Features →Automatic model generation, multi-model statistics, QAP, Mantel, GraphSAGE, and publication-ready exports.
See Analysis Tools →A method that reveals how people think about complex problems—not just what they think.
Fuzzy Cognitive Mapping (FCM) captures how people organize information in their heads. Instead of asking predetermined questions, you let stakeholders show you what concepts they think are important and how those concepts connect.
How it works: People identify the factors they think matter (like "pollution," "biodiversity," or "economic growth") and draw arrows showing how they influence each other. This creates a map of their thinking—their mental model—that can be analyzed mathematically.
Why it matters: Two people can have the same opinion but completely different reasoning. One person might see simple cause-and-effect. Another might see complex feedback loops. Traditional surveys can't capture this difference. FCM does.
Follow a study from concept design to AI-powered insights — no account required
Experts and stakeholders collaborate to identify the key concepts and causal relationships in the system. The model builder is used to sketch initial causal maps and build shared understanding.
Research Question
How do different stakeholder groups perceive the causal relationships driving the energy transition?
Through expert panels, literature review, and stakeholder input sessions, the research team identified 14 key concepts and mapped initial causal relationships. The single-model builder captures shared knowledge before deploying at scale.
Drag nodes to rearrange the layout
In the full platform: Run expert facilitation workshops, use the drag-and-drop model builder, and iterate on your study design before deploying.
The survey goes out to 847 stakeholders. Responses stream in and are automatically converted to individual mental models.
If Carbon Pricing Policy were to increase, how would it affect Economic Growth?
Not all items may be connected. Select “No Effect” or “N/A” if you believe there is no meaningful relationship.
Respondents rate the direction and strength of each causal relationship using a 7-point scale
Industry Executives response rate is 79.8%. Need more data? The platform can generate synthetic mental models conditioned on demographics using GAN-based predictive cognition to boost your effective sample size.
In the full platform: Build custom surveys with branching logic, deploy via email/SMS/QR codes/custom links, track responses in real-time with automated reminders, and quality control with built-in verification.
714 completed models are aggregated and compared. The platform reveals where stakeholders agree and where they fundamentally diverge.
Renewable Investment → Job Creation
94% of models (672/714), mean weight +0.72 (SD 0.08)
Carbon Pricing → CO2 Emissions
91% of models (650/714), mean weight -0.68 (SD 0.11)
Fossil Fuel → CO2 Emissions
89% of models (636/714), mean weight +0.74 (SD 0.09)
Carbon Pricing → Economic Growth
Scientists +0.45 vs Industry -0.62 (Kruskal-Wallis H=14.2, p=0.003)
Regulatory Framework → Innovation
Policymakers +0.55 vs Industry -0.38 (H=11.8, p=0.008)
Fossil Fuel → Energy Security
Public +0.71 vs Scientists -0.44 (H=16.1, p<0.001)
Scientists and Policymakers share the most similar mental models (r=0.71, p<0.001, large effect). Industry diverges most from the aggregate (r=0.34, p=0.02, medium effect). Based on 1,000 permutation tests.
Scientists' models average 31.2 edges (density 0.38) with 4.7 feedback loops. Public models average 18.4 edges (density 0.19) with 1.2 loops. Higher domain expertise correlates with more complex causal reasoning (Spearman rho=0.64, p<0.001).
95% CI across 999 bootstrap resamples. Most stable edge: Renewable Investment → Job Creation [0.67, 0.78]. Least stable: Fossil Fuel → Energy Security [-0.12, 0.53] — high disagreement across groups.
In the full platform: Run 25+ statistical analyses including Mantel tests, structural equivalence, motif detection, relationship discovery, Hedges' g effect sizes with Holm/FDR correction, split-half reliability, sample adequacy testing, and segment comparison across any demographic variable.
Go beyond descriptive statistics. AI and deep learning reveal hidden causal structure, optimal interventions, and predictive insights across the full dataset.
Adjust intervention strengths and see predicted system-wide impacts in real time.
AI Recommendation (Q-Learning)
Optimal policy: Carbon Pricing = 0.7, Renewable Investment = 0.8, Public Awareness = 0.6. Predicted impact 3.1x higher than baseline.
ML analysis identified 3 distinct stakeholder archetypes: Techno-Optimists (34%, emphasize Innovation → Growth), Policy Advocates (41%, emphasize Regulation → CO2 Reduction), and Status Quo Defenders (25%, emphasize Economic Growth → Energy Security). BIC: 1,247.
WGAN-GP generated 200 synthetic Industry Executive models conditioned on group demographics. Effective sample boosted from n=158 to n=358. Bootstrap CI for Industry edges narrowed by 41%. KL divergence between real and synthetic distributions: 0.08 (excellent fidelity).
In the full platform: Run GraphSAGE node embeddings, transformer attention analysis, GAN-based synthetic data generation, Monte Carlo sensitivity analysis, Bayesian uncertainty estimation, and more — all in your browser.
Complete research toolkit: Survey builder, data collection at scale, and research-grade analysis—all in one platform
Start with expert facilitation or design your own study
Expert-facilitated concept elicitation workshops. We help you design optimal FCM surveys by working with key stakeholders to identify critical concepts and relationships.
The only survey platform built for FCM data collection
"Complexity is an emergent property of aggregation, not a prerequisite for participation."
Participants contribute simple mental models. Complexity emerges when you aggregate hundreds or thousands of responses.
Build specialized FCM surveys with our drag-and-drop interface. Add concept lists, configure edge weights, and design the exact mental model capture you need. No other survey platform can do this.
Send surveys via email, SMS, or custom links. Embed on your website or share QR codes. Respondents can save progress and return later. Automated workflows trigger actions on completion, quality flags, or quota limits. Balanced concept coverage tracking, scheduled sends, and real-time response management. Scale from 10 to 1,000+ participants.
Each survey response is saved as an individual FCM and instantly combined into an aggregate model. No manual data entry, no Excel exports, no reformatting. No other platform does this—go from raw responses to publication-ready multi-model analysis instantly.
Research-grade tools for professional FCM analysis
Compare up to 1,500+ participant models simultaneously. QAP tests, Mantel analysis, edge frequency, structural equivalence, motif detection, relationship discovery, and concept selection analysis. Segment and compare how different stakeholder groups think with multi-level group comparison.
Split-half reliability, sample adequacy testing, formal meta-analysis with forest plots, consensus measurement, graph distance metrics, Hedges' g effect sizes with Holm/FDR correction, and Bayesian uncertainty with credible intervals on edge weights. 25+ analysis methods for research-grade results.
GAT and GraphSAGE models with attention mechanisms, transformer analysis, causal discovery, and GAN-based synthetic data generation. Discover hidden patterns and predict network behavior using cutting-edge deep learning.
Identify influential concepts with degree, betweenness, closeness, and eigenvector centrality. Find leverage points and key drivers in your system.
Run "what-if" simulations with FCM propagation. Clamp concepts to fixed values, iterate to equilibrium, and visualize how interventions cascade through the network. Compare baseline vs. intervention outcomes.
Drag-and-drop interface for creating FCM networks manually. Add nodes, draw weighted edges, and visualize causal relationships in real-time with interactive network layouts.
Designed for hundreds to thousands of respondents. Cultural consensus measurement, split-half reliability, power analysis, temporal tracking, multi-dataset merge with column harmonization, and per-group comparison across every analysis.
Find optimal intervention strategies using reinforcement learning. Q-learning identifies which concepts to activate or suppress for maximum impact on your target outcomes.
GAN-based behavioral prediction modeling. Generate synthetic stakeholder responses conditioned on demographics to predict how new populations would think about your problem.
Track how mental models evolve over time. Detect concept drift, structural change, and network evolution across multiple data collection periods.
Everything you need for professional research
Export high-resolution network visualizations, statistical reports, and interactive plots. Automated Word (.docx) and PDF report generation from analyses. CSV, Excel, SPSS, GraphML, PNG, and PDF formats supported. Ready for journals and presentations.
2FA (TOTP & email), WebAuthn/passkey login, SSO (SAML 2.0 & OIDC), AES-256 encryption, CSRF protection, intrusion detection, and automated security scanning. GDPR compliant with full data export and deletion.
FCM is used across industries to solve complex problems and make better decisions
Model climate change impacts, ecosystem dynamics, and conservation strategies. Analyze stakeholder perspectives on natural resource management and identify leverage points for policy intervention.
Map disease transmission pathways, healthcare system dynamics, and social determinants of health. Compare stakeholder mental models to identify intervention strategies.
Evaluate policy impacts across multiple dimensions. Capture diverse stakeholder perspectives to build consensus and identify unintended consequences before implementation.
Model organizational dynamics, strategic risks, and decision-making processes. Compare executive perspectives to align understanding of business challenges.
Analyze fishing satisfaction, habitat threats, and conservation solutions. Model stakeholder knowledge to inform sustainable fisheries management.
Conduct participatory research with publication-ready visualizations. Compare mental models across demographics using rigorous statistical methods.
Get answers to common questions about Get Causality
We measure how people organize information in their heads—their mental models. Surveys tell you people's opinions. We show you how they're thinking about the problem. Do they see it as one thing causing another in a line? Or do they see a complex web of interconnected causes? That difference matters because you can't convince someone by arguing against their mental structure.
A mental model is how someone internalizes cause-and-effect relationships in their head. It's their personal theory of how things work. For example, one person might think "more regulation → less pollution → better health." Another might think "more regulation → higher costs → job losses → worse health." Same topic, completely different mental models. Understanding these differences is key to effective communication and policy design.
Three things: (1) What concepts people connect—Do they link "climate change" to "fishing" or not? (2) How they connect them—Does A cause B, or do they affect each other? (3) How complex their thinking is—Simple chains vs. feedback loops. The output is a map of their thinking, plus statistics showing if different groups think fundamentally differently.
(1) Sign up—free tier available, no credit card required. (2) Create a survey using our drag-and-drop builder, or import existing data. (3) Share a link—participants draw their mental models in their browser (no app install). (4) Platform aggregates responses into analyzable networks automatically. (5) Run analyses with one click—centrality, clustering, group comparisons. (6) Export publication-ready visualizations and statistics.
Anyone who needs to understand stakeholders before making big decisions: consultants doing stakeholder analysis for clients, companies launching new products or policies, governments designing public policy, and researchers studying how people understand complex issues. All of them currently guess at how stakeholders think. We measure it.
Surveys ask questions you think are important. We let people tell us what they think is important and how it all connects. Example: A survey asks "Do you support climate policy?" and gets 60% saying yes. Our platform shows you why they support it—some see economic benefits, others see moral duty, others see system collapse. Same opinion, completely different thinking. You need to know the why to design messages that work.
Traditional FCM analysis can take weeks between getting survey responses and having results—most of that time spent on data cleaning, reformatting, and writing R scripts. Get Causality automates this workflow. Import your data, select your analyses, and get results in minutes instead of weeks. The platform is designed to handle thousands of participant models with demographic comparisons and automated visualizations.
The free plan lets you work with up to 250 FCM models and send up to 50 survey invitations. You can import and export data, visualize networks, and run basic centrality analyses. It's perfect for students, small pilot projects, or just testing the platform to see if it fits your needs. No credit card required.
We support most common FCM data formats. You can import survey exports (CSV or JSON), standard FCM JSON files, Excel adjacency matrices, or just paste directly from a spreadsheet. The platform automatically detects your data format and handles the conversion, so you don't need to manually reformat anything.
25+ analysis methods across two modules. Multi-Model: QAP tests, permutation tests, edge frequency, structural equivalence, motif detection, network dynamics, relationship discovery, concept selection analysis, bootstrap confidence intervals, loop influence analysis, and segment analysis for comparing stakeholder groups. Meta-Model: TCEC, Monte Carlo, Bayesian uncertainty, concept preference, discrete choice modeling, ML ensemble, SHAP explainability, network meta-analysis, latent class, GraphSAGE, causal discovery, RL-based intervention optimization, predictive cognition, and scenario modeling. Population-scale: split-half reliability, sample adequacy, consensus measurement, graph distance metrics, formal meta-analysis with forest plots, intervention propagation simulation, temporal analysis, and multi-dataset merge for cross-study comparison. All with automated visualizations, interpretation guides, automated Word/PDF reports, and exportable results.
Yes. The platform runs on a serverless architecture (Vercel + Neon PostgreSQL) with no persistent local storage—each function is stateless and isolated, so there is no shared server disk where data could be exposed. All data transmission is encrypted via TLS, sensitive fields are encrypted at rest with AES-256-GCM, and passwords are hashed with bcrypt. We enforce strict Content Security Policy headers, CSRF protection, HSTS, and rate limiting across all endpoints. A built-in PII filter automatically strips personally identifiable columns (names, emails, IP addresses, phone numbers, government IDs) from uploaded data before it reaches any analysis pipeline. Two-factor authentication is available via TOTP authenticator apps or email codes. Enterprise users get SSO (SAML 2.0 & OIDC) and per-institution 2FA policies. All users can configure personal IP allowlists to restrict login by network. Our CI/CD pipeline runs automated security scans (dependency audit, secrets detection, SAST) on every deployment. GDPR compliant with full data export and account deletion. If you have specific IRB or institutional requirements, contact us to discuss custom data handling agreements.
Enterprise and Team tiers include institutional SSO (SAML 2.0 and OpenID Connect) with email domain auto-detection, per-institution 2FA enforcement, trusted device management, comprehensive audit logging for compliance, intrusion detection alerts for suspicious activity, and configurable session policies. All users can set up personal IP allowlists to restrict login to trusted networks. We also support WebAuthn/FIDO2 for passwordless authentication. The entire backend runs on Vercel’s serverless infrastructure with Neon PostgreSQL—there are no persistent servers to compromise, and each function executes in an isolated, stateless environment. Sensitive data columns are automatically excluded from analysis pipelines via client-side PII filtering before data ever leaves the browser. All security events are logged with user ID, IP, and timestamp for IRB and compliance review.
You can pay monthly (cancel anytime) or annually (save 15-20%). All plans include feature updates as we add them. Beta users who join now can lock in founding member pricing at up to 50% off for life. If you have a team of 10 or more, contact us about enterprise pricing.
Absolutely. The free plan gives you full access to core features with no credit card required. You can also try the live demo (check the "Live Demo" tab) to explore with example models. Or join the beta waitlist for early access with founding member pricing. Take your time and see if it fits your workflow.
Get Causality is currently self-funded and bootstrapped. Revenue comes from platform subscriptions (individual, team, and enterprise tiers) and professional workshop services. We are actively exploring strategic partnerships and investment to accelerate product development, expand our research network, and scale enterprise adoption.
Yes. We are open to conversations with investors who understand the value of decision intelligence, stakeholder analytics, and research infrastructure. If you're interested in learning more about our growth trajectory, market opportunity, or partnership possibilities, please reach out to [email protected].
Just reach out. For general questions, email us at [email protected]. If you want to schedule a demo for your team, contact [email protected]. Paid users get technical support with a 24-hour response time. Beta users get priority support and direct access to our development team.
Choose the right solution for your research needs
Choose your level of support: Self-service platform access includes survey builder + analysis suite. Add workshop services for expert-guided concept design. Bundle for complete end-to-end support.
Perfect for learning FCM basics
Requires .edu verification
For educational use only
Multi-Model + Meta Models
For educational use only
Unlock Predictive Cognition
For educational use only
Commercial license for consultants
Full platform + commercial license
For research teams & organizations
Reach out for pricing
All paid plans include: Unlimited responses per survey and unlimited participant models. Survey storage limits vary by plan. Plans differ by which analyses are available, license type, and team features.
Facilitated workshops for exploration, training, and strategic engagement
Half-day facilitated session
1.5-2 day workshop
3-5 workshops over 2-6 months
Long-term strategic partnership
All workshops include: Expert facilitation, comprehensive deliverables (10-30 pages), and methodology training. No excessive reports.
Add-on services to enhance your survey success (requires active SaaS subscription)
Expert survey design review
Network assessment & planning
Boost completion rates
Complete survey support
Note: Survey support services require an active Academic, Professional, or Team subscription. You handle survey distribution through your own networks.
Comprehensive packages combining platform access, workshops, and support services
Perfect for starting your first project
For comprehensive strategic projects
Long-term strategic partnership
All bundles include: Expert facilitation, comprehensive training, ongoing support, and significant cost savings. Custom bundles available.
| Feature | Free | Academic | Acad Plus | Acad Pro | Researcher | Professional | Team | Enterprise |
|---|---|---|---|---|---|---|---|---|
| Monthly Price | $0 | $249 | $349 | $449 | $499 | $999 | $1,999 | Custom |
| Basic Analysis | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Multi-Model Analyses (QAP, Mantel, GNN, ML, Cluster, Motif, Discrete Choice, Latent Class) | — | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Meta Model Analyses (25+ AI/ML + Causal Discovery) | — | — | ✓ | ✓ | — | ✓ | ✓ | ✓ |
| Predictive Cognition (GAN, Transformer, GraphSAGE) | — | — | — | ✓ | — | ✓ | ✓ | ✓ |
| Goal-Seek Optimizer | — | Scan + Optimize + Compare | All modes | All modes | Scan + Optimize + Compare | All modes | All modes | All modes |
| Stored Surveys | 1 | 10 | 25 | 50 | 10 | 50 | 100 | Unlimited |
| Responses per Survey | 100 | Unlimited | Unlimited | Unlimited | Unlimited | Unlimited | Unlimited | Unlimited |
| A/B Testing & White-label Surveys | — | — | — | — | — | ✓ | ✓ | ✓ |
| Team Members | 1 | 1 | 1 | 1 | 1 | 3 | 10 | Unlimited |
| Commercial License | — | Edu only | Edu only | Edu only | ✓ | ✓ | ✓ | ✓ |
| SSO / SAML | — | — | — | — | — | — | — | ✓ |
| Support | Community | Email 48h | Email 24h | Priority 24h | Priority 24h | Email 24h | Priority 12h | Dedicated |
Yes, you can upgrade to annual billing anytime to save up to 20%.
Each plan has a cap on how many surveys you can store at once (e.g., Academic = 10, Team = 100). Responses per survey are unlimited on all paid plans. If you need more survey slots, upgrade your plan or archive old surveys to free up space.
They unlock the same analyses, but Academic requires .edu verification and is for educational use only. Researcher includes a commercial license for consultants and independent researchers.
Yes! Our Academic tiers are specifically priced for students and faculty. Contact us for additional institutional discounts.
Enterprise includes everything in Team plus SSO/SAML, custom domains, dedicated support, and custom integrations. Reach out to discuss your needs.
Have questions? We'd love to hear from you.
See how Get Causality transforms data into insights