02. Market Analysis

Chapter 2 of 24 · 15 min

KEY INSIGHT

The addressable market for local AI products isn't everyone with internet access—it's the intersection of people with problems AI solves, willingness to pay, and access to mobile money or bank transfers. Start with segmentation. Nigerian AI consumers break into rough categories: English-speaking professionals earning ₦500,000+ monthly who already use ChatGPT (high willingness, low urgency for local alternatives), Pidgin speakers and low-English proficiency users locked out of global AI tools (low willingness, high need), and businesses seeking automation (high willingness, complex implementation). The business segment is underserved but expensive to acquire. The low-English-proficiency segment has genuine need but payment friction. Your product positioning depends on which segment you're building for. Secondary data sources that actually work in Nigeria: - NITDA reports and National Bureau of Statistics tech surveys - GSM Association mobile money reports for payment infrastructure data - Fintech operator disclosures (Flutterwave, Paystack annual reports) - Operator partnerships and informal data sharing with local AI communities Avoid over-reliance on global AI market reports. They systematically underestimate African markets because they model on US/European adoption curves and don't account for leapfrog dynamics. ```python import pandas as pd import numpy as np def segment_market_size( population, internet_penetration=0.55, smartphone_rate=0.70, target_segment_pct=0.15, willingness_to_pay=3000, serviceable_pct=0.30 ): """ Bottom-up market sizing for Nigerian AI product. All monetary values in naira. """ # Start with population, apply filters online_population = population * internet_penetration smartphone_users = online_population * smartphone_rate target_segment = smartphone_users * target_segment_pct # Serviceable addressable market (users you'd realistically reach) serviceable = target_segment * serviceable_pct # Potential annual revenue annual_revenue = serviceable * willingness_to_pay * 12 return { 'total_population': population, 'online_population': online_population, 'smartphone_users': smartphone_users, 'target_segment': target_segment, 'serviceable_market': serviceable, 'annual_revenue_potential': annual_revenue, 'monthly_revenue_potential': annual_revenue / 12 } # Lagos metro area example lagos = segment_market_size(15_000_000) # Greater Lagos print(f"Serviceable market: {lagos['serviceable_market']:,.0f} users") print(f"Monthly revenue potential: ₦{lagos['monthly_revenue_potential']:,.0f}") ``` Real failure mode: Market analysis paralysis. You'll never have perfect data. Launch a landing page, collect emails, charge actual money. The market tells you what it wants through conversions, not spreadsheets.

Market analysis for local AI products requires separating signal from noise in data that doesn't exist yet. Nigeria's AI market lacks reliable public datasets. You're building analysis from operator experience, industry reports, and direct observation.

EXERCISE

Define your three primary user segments. For each, identify: the problem you solve, their payment method, and one friction point that would cause churn.