1 A hypothetical example of mock data - demand for Greek yogurt

Projection of demand for Greek yogurt

  • 20 countries (5 randomly selected from 4 sub-regions)
  • 2024-2036
  • Driven by a few factors - i.e., assumptions

Example of potential assumptions

  • Market increase: market size change for overall dairy and/or health food (constant/stable rate of change) - 5 scenarios
  • Greek-yogurt specific demand change (constant/stable rate of change) - 3 scenarios (no, slow, fast increase) expressed in country-specific numeric values
  • Tax on dairy food (shock) - not included in the example
  • Real/Fake news about the product (shock) - not included in the example

In this example, users can change two assumptions.

1.1 Data schema

For the app, we need only

  • country
  • region
  • year
  • [USER INPUT] columns (i.e., select assumptions)
  • [OUTPUT] demand
Column name Example values Type Number of unique values Range Note
country Kenya, Madagascar, Rwanda character 20 Country
region Eastern Africa, Southern Africa, Southern Asia character 4 Region
year 2024, 2025, 2026 integer 12 2024 – 2035 Year
scenario_id Fast increase: market projection 97.5th percentile, Fast increase: market projection 90th percentile, Fast increase: market projection 50th percentile factor 15 Scenario ID for internal purposes
market_direction Upper, Median, Lower character 3 Market size projection probability direction
market_projection 97.5, 90, 50 numeric 5 2.5 – 97.5 [USER INPUT 1] Market size projection probability
market_size 56565, 57957, 59400 numeric 1153 474 – 206423 Market size
growth_scenario Fast increase, Slow increase, No change factor 3 [USER INPUT 2] Product growth scenario
base_rate 20, 30, 10 numeric 3 10 – 30 Base rate of change for product growth
annual_change 1.5, 0.5, 0 numeric 5 0 – 3 [USER INPUT 2] Annual rate of change for product growth, country-specific numeric value per growth scenario
growth_rate 0.2, 0.215, 0.23 numeric 47 0.1 – 0.43 Product growth rate
demand 11313, 12461, 13662 numeric 2766 142 – 84633 [OUTPUT] Demand for the product (in thousands)

How much and what information do you want to share with users?

  • market size is an important intermediary outcome, but it can be information overload.
  • annual rate of growth can be country-specific numeric (annual_change) or categorical (growth_scenario).

1.1.1 FYI, further look into growth scenarios

Currently same within a region, just for the purpose of this exercise. This will be likely country specific in the actual DMPA-SC data (mock or real)…

Region Current use rate (base_rate) (%) APPC no change APPC slow increase APPC fast increase
Western Africa 10 0 1.0 3.0
Southern Asia 30 0 0.5 1.0
Eastern Africa 20 0 0.5 1.5
Southern Africa 20 0 0.5 1.5

APPC: Annual percent point change

1.2 Projection by scenario - country/region examples

1.2.1 Togo, Western Africa

1.2.2 Kenya, Eastern Africa

1.2.3 Lesotho, Southern Africa

1.2.4 Sri Lanka, Southern Asia