Updated 14 October 2026: expanded with practical context and refreshed information. Original publication date retained.
Township size depends on the census year and boundary used. A name may refer to a main place, several wards or a broader urban area. Business planning such as a spaza-shop model should not use an invented current population as market evidence.
Start with the statistical geography
Stats SA’s Tembisa main-place profile labels its key figures as 2011 and records 463,109 residents. That is a historical count, not a verified 2025 or 2026 estimate.
The 2022 ward-level small-area estimate release describes a different product. A ward total should not silently replace a main-place total without checking the geographic relationship.
Major areas to research
Soweto, Tembisa, Katlehong, Umlazi, Soshanguve, Khayelitsha, Mamelodi, Mitchells Plain, Ibhayi and Sebokeng are examples commonly discussed in this context. This is a research shortlist, not a certified ranking by current population or area.
A rigorous table uses the same year and boundary type for all entries. If combining subplaces, document the components and avoid counting neighbouring settlements twice. Our Maponya Mall history concerns a specific Soweto development, not a census of the whole area.
People, households and density
Population counts people; household counts describe households; density relates people to land area. These measures cannot be substituted for one another. A large place can have a different density from a smaller, more concentrated one.
National housing percentages should not be copied into every local profile. Use the place’s own statistics and retain their year rather than presenting historical conditions as newly measured facts.
Apply the data carefully
Catchment boundaries, competitors, transport and the product market also matter. Our road-freight guide provides a distribution context without establishing local purchasing power.
For current data access, use the Stats SA census portal. This guide avoids fabricated future projections and explains how a defensible comparison can be made from consistent geographic records.
Translate a township profile into a realistic customer catchment
Stats SA's census resources provide the starting point for population data, but a business catchment is not automatically the same as a township's statistical boundary. The existing 2011 main-place figures and newer ward products need their own geographic labels before they can support a comparison.
For a shop, map the area customers can reach conveniently, the routes they use and competing services. A household near the boundary may buy elsewhere, while a commuter from outside may shop near a transport stop. Total residents alone do not establish the number of paying customers.
Collect a small amount of local trading evidence alongside the census context. Record enquiries, transaction times and which products customers repeatedly request. Avoid replacing an old population count with an invented current growth estimate merely to make the market look larger.
Use one geography and year consistently in a ranking, and document any aggregation. For the individual venture, connect that evidence with observed demand and delivery costs. This adds an operating application to an already improved statistical guide without reintroducing unsupported population projections.
Source and further reading: Translate a township profile into a realistic customer catchment — reference.













