Hi there,
complementing Allen - I would say the logic building blocks in real cases are similar vs interview
e.g. I had to size the solar pv, energy efficiency market evolution at 5 years for specific residential and industrial segments
We built a semi-sophisticated excel model but basically my team
1) estimated “smart”drivers and proxies for market size evolution based on reports, business sense, client knowledge, bcg knowledge
→ you will size most of the time obscure sub-segments of market where no deterministic data points are available
2) triangulated / pressure-tested outcomes
in summary a more in depth exercise vs an interview but built on EXACTLY the same conceptual tools and reasoning (hence the relevance of asking mkt sizing questions during interview :) )