Enobl VenturesWhat is being built?
Business description, maturity, age, industry, geography and operating status.
EV SCORE
Proprietary intelligence.
Human conviction.
EV Score is our proprietary predictive system for evaluating early-stage companies. It draws on company, founder, financing, and market information to rank opportunities and guide our diligence.
01 FROM SCORE TO SELECTION
A clearer focus
for deeper diligence.
We focus on the top 25% of companies by EV Score. Our investment team then evaluates the business and technology before deciding whether to invest. A high score is not a guarantee of future performance.
02 THE SIGNALS
A company in context.
The system brings together multiple dimensions of a company, rather than relying on a pitch or one isolated metric.
Who is building it?
Founder background, biography, education and the context around the team.
How has it developed?
Funding history, round sequence, prior investors, valuations and time between rounds.
What surrounds it?
Overlapping sector classifications and the wider market and economic environment.
Signal categories drawn from the system’s documented development. Available information varies by company.
03 THE MODELING FOUNDATION
Built on years of
venture modeling.
Much of a startup’s story lives in words. Our modeling approach brings descriptions and biographies together with financing and company data, drawing on more than a decade of development and refinement.
Turn information into comparable signals.
Data is cleaned and normalized. Natural-language processing uses word embeddings to turn company descriptions and founder biographies into numerical representations, helping models recognize similarities even when companies describe themselves differently.
Our work with feature reduction has included PCA, PLS, UMAP-related approaches, and neural-network autoencoders. These techniques create more compact representations of complex data, aiming to retain useful patterns while reducing redundancy. They help the models work with the information behind the words.
Ask several questions of the same company.
Classification and regression models address different questions. Classification models identify categories or estimate the likelihood of events, such as further funding or an exit. Regression models estimate numerical values, such as a valuation at a future financing round.
Models tailored to a company’s stage, together with ensemble approaches that combine several models, bring multiple perspectives to the evaluation. The methods used depend on the question and the available data.
Test on information the model hasn’t learned from.
We evaluate predictions on data held outside model training to test how well the models generalize beyond what they have already seen. These out-of-sample results help us assess performance and calibrate estimates of uncertainty.
We continue to invest in research and refinement as data and markets change. Our investors and operators bring the context models can miss: the people, the technology, and what is happening in the world. The score informs their investigation; it does not make the investment decision.
04 TESTING THE METHODOLOGY
Predictions recorded
before outcomes were known.
In November 2022, approximately 40,000 companies were scored. Subsequent analysis found that higher-scoring companies were more likely to secure funding or achieve an exit.
The study examined company outcomes, not investment returns. It does not establish future fund performance.
05 UNCERTAINTY IS PART OF THE WORK
A range of possibilities.
A more informed question.
The valuation methodology was designed around distributions of possible values at a future financing event. That makes uncertainty visible rather than hiding it inside a single estimate.
Valuation modeling is one part of the broader architecture. A modeled financing valuation is not an exit prediction, and a score does not guarantee an outcome.
06 THE INVESTMENT DECISION
Evidence to investigate.
Judgment to act.
Our models help prioritize attention and inform questions. Experienced investors and operators conduct diligence and make every final investment decision.