An increasing amount of value – both at the enterprise and societal level – is held as software or data. But while we can acknowledge that software algorithms, machine learning datasets, and real-time data feeds have value, who can say how much value?
Historically, software and data assets and companies have been valued using models like the Constructive Cost Model (COCOMO). These models often estimate the effort required to reproduce a certain number of lines of code, then convert this effort into a monetary value using labor costs and programmer productivity assumptions.
As famously attributed to George Box: “All models are wrong but some are useful.” COCOMO has been a useful model historically to help provide high-level, order-of-magnitude estimates. Is this project’s replacement cost closer to $100,000 or $100,000,000? However, in the M&A or insurance context, there’s a big difference between $10M and $50M. Don’t you think you’d care if you were buying or selling a company? We have, and we still do – which is why we’ve developed advanced models for software valuation and data valuation.