Every market-entry plan, investment memo and annual plan starts with a market size. Most go wrong not because of bad arithmetic but because of unchecked assumptions.
Top-down
Start from a broad number, such as total households, total retail spend or category consumption, then apply filters: share that is urban, share that is online, share that buys the category. It is quick and good for orders of magnitude, but every filter is an assumption that compounds.
Bottom-up
Build from units you can observe: number of buyers × purchase frequency × spend per purchase, or number of stores × throughput × ticket size. It takes longer, but each input can be checked through surveys, channel checks or expert calls.
Triangulate
A robust estimate uses both and reconciles them. If they differ widely, the gap tells you which assumption to investigate. A supply-side cross-check (what producers and distributors say they ship) is a valuable third angle.
Common pitfalls
- Confusing total addressable market with what is actually serviceable today.
- Applying metro behaviour to Tier-2 and Tier-3 cities.
- Mixing value at MRP with value at selling price.
- Reporting false precision. A clear range with stated assumptions beats a single unsupported number.
What a good sizing deliverable looks like
A range, a transparent assumption tree, sensitivity to the two or three inputs that matter most, and a note on how to update the estimate as new data arrives.