Tilda is built on decades of research across disciplines, including the largest-ever meta-analysis of behavior change, as well as foundational insights from psychology, neuroscience, learning science, and organizational theory. Developed in collaboration with researchers at RISE Research Institutes of Sweden, Tilda helps teams design initiatives that lead to real change. Grounded in science, made for the real world.

‍

Our Mission: Bridging Science and Practice

We started Tilda out of frustration. Despite all we know from decades of research, 70–90% of learning and change efforts still don’t lead to meaningful results. The science is there, in academic papers, models, and frameworks, but rarely does it reach the people who need it most, in a form they can actually use.

Tilda exists to bridge that gap.

To take the best from science and turn it into tools that are simple, practical, and powerful. And over time, we hope to do the reverse too, helping practice feedback into research.

‍

Our Foundation: A Scientific Model for Practical Change

One of the core foundations behind Tilda is the 2024 meta-analysis by Albarracín et al., the most comprehensive study to date on behavior change, analyzing over 2,600 effect estimates across 100+ behavioral determinants. It gives us one of the clearest pictures yet of what actually works.

Together with our research partners at RISE, we’ve built a model that combines insights from over 40 foundational studies across fields like:

• Social psychology
• Behavioral science
• Neuroscience
• Learning and cognitive science
• Organizational psychology
  
  

Our model is explicitly two-level:

• It captures individual drivers like skills, confidence, emotion, and motivation
  
• And structural factors like norms, access, tools, and incentives
  
  

Applying the Science: Tilda’s Behavior Change Score

Designing for change means knowing what to expect from your efforts and how to improve them. That’s what the Behavior Change Score™ is for.

Behind the scenes, Tilda analyzes every activity in your initiative and maps it to the behavioral drivers it targets, from high-impact factors like practice and purpose to often-overrated ones like general knowledge. Each factor is weighted by how much it actually influences behavior, based on published effect sizes.

Then, we estimate the overall likelihood your initiative will succeed and show where you can make the biggest improvements.

ChatGPT can explain behavior science. Tilda can quantify it. We combine conversational AI with a causal engine that knows what drives change and how strongly.

Unlike traditional tools that give you templates or advice, Tilda helps you understand the effectiveness of your own design, using real science, every step of the way.

‍

Where We Are and What’s Next

Tilda’s prediction engine is already live in a public beta, used by real teams to design smarter, faster, and with more confidence. Together with researchers, we’re continuing to evolve the model, building a system that not only reflects the best available evidence, but keeps improving as more real-world data flows in.

In partnership with RISE, we’re exploring advanced approaches like Bayesian modeling and structural causal graphs to better capture combination effects, uncertainty, and variation across different audiences and behaviors.

Our goal:

To make it radically easier to design initiatives that actually work. And to keep raising the bar for what "good design" means in learning and change.

‍

Research References

Albarracín, D., Fayaz-Farkhad, B., & Granados Samayoa, J. A. (2024). Nature Reviews Psychology, 3(6), 377–392.

Ajzen, I. (1991). The Theory of Planned Behavior.
Bandura, A. (1977). Self-Efficacy: Toward a Unifying Theory of Behavioral Change.

Deci, E. L., & Ryan, R. M. (2000). Psychological Inquiry, 11(4), 227–268.

Ericsson, K. A., et al. (1993). Psychological Review, 100(3), 363–406.

Salas, E., et al. (2012). Psychological Science in the Public Interest, 13(2), 74–101.

Cialdini, R. B., & Goldstein, N. J. (2004). Annual Review of Psychology, 55, 591–621.

RISE Research Institutes of Sweden (2024–). Model development partner for Tilda’s Behavior Change Score.‍

Full reference list available upon request.