REGIONAL VARIATION IN STARTUP VALUATIONS AND
SURVIVABILITY
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Startup valuations vary a lot across regions/industries even
with similar economies & investors. Traditional models don't
capture this well.
Segmented models used more by practitioners than
academia. Allow qualitative factors like management,
environment, business model. Provide flexibility.
Examples: Berkus, Payne, Kauffman scorecards. Weight
factors like market potential, IP, team experience.
Peer reviewed models also segment by financials,
operations, deal terms. Sum factors.
Hierarchical Microtargeting
Machine learning enables more sophisticated segmentation
via decision trees and random forests.
Microtargeting inductively analyzes granular data patterns
like geography, industry, model. Reveals complex
interactions missed by regressions.
Partitioning creates homogeneous valuation subsets. Shows
where/how much each factor matters.
Applications Beyond Valuation
Also useful for modeling startup selection and survivability.
Selection models often qualitative criteria weighted by
importance.
Survivability also shows regional/industry variation. Decision
trees could enrich regressions with macro factors.
Key Takeaways
Allow qualitative factors critical for opaque startups
Granular insights into interactions
Strengthen selection/survivability modeling
Underutilized in academic literature vs practitioners
