CulturePulse
Predict Human Behavior
Lightweight
Cost-Efficient
Psychographic Mapping
September, 2021
All commercial information and intellectual property contained in this document are strictly confidential for the client, its employees, and its collaborators.
Overview
If you know someone’s beliefs, you can predict their behavior.
We predict human behaviors with quick and affordable algorithms that map beliefs and provide more actionable market insights.
Problem
Companies are making decisions by counting likes and views but can’t access the beliefs that drive those customer behaviors.
Popular AI models (e.g., BERT and GPT) are costly to deploy and require constant retraining. Historical analysis fails to predict emerging trends and unprecedented events because the training data does not yet exist.
The market lacks quick, flexible, and affordable systems �for predicting how humans will receive new information.
Opportunity
More data doesn’t guarantee better insights.�Better algorithms guarantee better insights.*
CulturePulse has developed algorithms that map �beliefs and values and enable the prediction of human behavior.
These psychographic maps don’t require constant retraining �and work with any text in over 40 languages.
*Greco, Polonioli, & Tagliabue (2019). “Less (Data) Is More: Why Small Data Holds the Key to the Future of Artificial Intelligence“. https://arxiv.org/abs/1907.10424
Solution
We compare how a message fits with a market by profiling the market, modeling belief systems, and simulating thousands of responses.
We translate how the message fits the market into a ‘Resonance Score’: a single actionable metric for measuring the success of a message.
Based in evolutionary psychology, not machine learning, the resonance score is a metric for finding audiences where a message resonates the most and finding the audience that is more likely to share the message.
* The CulturePulse system is 130mb. BERT is > 1.3GB and can be > 4GB when trained. GPT2’s deployment is >5Gb.
Technology
Developed from 10 years of lab research and provides accuracy levels that are acceptable in clinical research settings (>95% reliable correlation).
Proven to provide deeper insights in multiple industries including social, finance, entertainment, security and defense, government, etc. over 4 years of consulting (revenue generated and reinvested in development).
* The CulturePulse system is 130mb. BERT is > 1.3GB and can be > 4GB when trained. GPT2’s deployment is >5Gb.
** See full taxonomy for all categories in CulturePulse’s ontology on Slide 25
*** This is also true for all word embedding systems, which do not have definable representations for specific elements in the array
Business Model - SaaS
Consumer Products
Business Products
API Licencing
*Domains include personality, morality, sentiment, social concerns, etc. See slide 32 in Appendix for full list of dimensions
Use of Funds
We’re raising a $500K SAFE with three goals:�
Financials
Case Studies
#1 Electric Vehicle Consumer Concerns
Problem
EV adoption is lagging in the US compared to the EU and consumer concerns are hard to track. Nationally representative surveys are expensive and time-consuming.
Strategy
Track conversations related to EVs on social media, and query a digital twin with key topics to see what will be most concerning to the market.
Results
CulturePulse was able to replicate market concerns found in nationally representative surveys. Results were provided in hours (not days).
Key Market Concern | Our Predicted Tweet Rank | % Ranking Concern as Most Important (Survey) |
Cost/Price premium | 1 | 26% |
Driving range | 2 | 24% |
Lack of charging infrastructure | 3 | 22% |
Time required to charge | 4 | 10% |
#2 COVID-19: Norwegian Government
Problem
Emotional contagion and misinformation were spreading quickly in Scandinavia related to COVID-19.
Strategy
Track 24 COVID-19 related hashtags using CulturePulse’s social listening system to understand what makes COVID-related tweets successful.
Results
Increased strategic knowledge for Norwegian ministry of health related to the emotional signatures of successful information and what kinds of topics the population was concerned with. Helped shape communications in later waves of the pandemic. Able to create a digital twin of the Scandanavian Twittersphere to replicate real-world results.
#3 Market Fit For Scripts
Problem
Entertainment marketing agency wanted to predict a script’s performance in specific markets.
Strategy
Utilize CulturePulse for testing messages to see if the script is compatible with Netflix audiences.
Results
Found that in key areas related to cultural beliefs and values, the script resonated with key subsets of the Netflix user-based (from Reddit data). Script sold to Netflix for USD20m.
#4 Market Fit For Scripts
Problem
Entertainment marketing agency wanted to test script compatibility with leading Hollywood actor.
Strategy
Utilize CulturePulse for testing messages to see if the script is compatible with public perception of the actor.
Results
Found that the top choice for the actor was not the best choice for script fit. Movie moving to pre-production and casting.
The Team
Justin E. Lane
D.Phil, University of Oxford
CEO, Co-Founder
F. LeRon Shults
2x Ph.D, Princeton
CRO, Co-Founder
Brian Swichkow
7 years as Growth Hacker
CMO
Maja Roknic
7 years as PM
Lead Product Manager
Advisors
John Zdanowski
Former CFO at Second Life, General Assembly, Assembled Brands, etc.
Jim Goodman
Creator of Ticketmaster Online and Tickets.com, formerly with Disney and Intel
Will Cady
Global Director of KarmaLab at Reddit
“No one ever made a decision because of a number. They need a story.”
― Daniel Kahneman
(2002 Nobel Prize Winner: Economics)