Fermi Estimation & Guesstimates: Formula Cheat Sheet
Guesstimates (or Fermi problems) test a candidate’s structured logic, comfort with numbers, and sanity-checking capabilities. Success depends on breaking a complex query down into a series of solvable, logical assumptions, rather than memorizing exact data.
Indian Demographic Reference Table
When estimating market demand, volumes, or capacity in India, use these standard demographic benchmarks to anchor your assumptions:
| Demographic Variable | Value / Percentage | Application / Rule of Thumb |
|---|---|---|
| Total Population | 1.4 Billion (1,400 Million) | Base population for country-wide estimations |
| Urban / Rural Split | • Urban: $35%$ (~500M) • Rural: $65%$ (~900M) | Critical for tech, retail, and FMCG penetration |
| Average Household Size | • Urban: 4 members / household • Rural: 5 members / household • Average: 4.5 members | Dividing population to find number of households (~300M total) |
| Income Tiers (Urban) | • High Income: $10%$ (~50M) • Middle Income: $40%$ (~200M) • Low Income: $50%$ (~250M) | Segmenting luxury, mid-range, and budget services |
| Age Group Structure | • $0 - 14$ years: $25%$ (~350M) • $15 - 64$ years (Working): $65%$ (~910M) • $65+$ years: $10%$ (~140M) | Useful for schools, employment, or pension-related cases |
| Gender Ratio | • Male: $52%$ (~730M) • Female: $48%$ (~670M) | Segmenting gender-specific products |
| Tech Penetration | • Internet Users: ~60% (~840M) • Smartphone Users: ~45% (~630M) | Base for e-commerce, digital payments, and app usage |
Core Formulas for Estimation
Use these foundational equations depending on the problem structure:
1. Demand-Side (Top-Down) Formulation
$$\text{Annual Demand} = \text{Total Population} \times \text{Penetration Rate} \times \frac{365}{\text{Replacement Cycle (Days)}} \times \text{Units per Event}$$
2. Supply-Side (Bottom-Up / Capacity) Formulation
$$\text{Daily Capacity} = \text{Number of Units} \times \text{Hours Operational} \times \text{Utilization Rate} \times \text{Throughput per Hour}$$
3. Little’s Law (Queueing / Retail Flow)
$$\text{Average Number of Customers (L)} = \text{Arrival Rate } (\lambda) \times \text{Average Time Spent (W)}$$
Common Guesstimate Pitfalls & Errors
[!WARNING]
- Double Counting: Adding “office workers who drink coffee” and “commuters who drink coffee” without subtracting the overlap.
- Ignoring Bottlenecks: Estimating that a toll booth can process 60 cars a minute when it takes 15 seconds to pay (maximum throughput is 4 cars per booth per minute).
- Failing to Sanity Check: If your guesstimate of the Delhi coffee market exceeds total Starbucks global revenue, stop and check your decimal points.
Step-by-Step Execution Playbook
- Clarify Scope: Ask if “daily coffee cups in Mumbai” includes instant coffee at home or only commercial cafes.
- Select the Model: Choose between Demand-Side (e.g., how many people want to buy coffee) and Supply-Side (e.g., how much coffee Mumbai’s cafes and vendors can produce).
- Execute Math with Round Numbers: Use $1.4\text{B}$ instead of $1.428\text{B}$. Keep numbers clean to avoid mental arithmetic errors.
- Sanity Check: Compare your final result to a known baseline (e.g., India’s GDP, car ownership percentages, or internet base).
[!TIP] Always state your assumptions out loud. Interviewers care about the process of how you got to the number, not the actual number itself. Say: “I am assuming that a middle-class urban household owns 1.5 two-wheelers on average.”
To practice interactive guesstimates and consulting math under real-time pressure, visit caseedge.in.