Thetha AI
The sector, honestly
Everything below is drawn from public research and government reporting. Where sources disagree — and on this subject they frequently do — the disagreement is stated rather than resolved.
How big is it? Nobody actually knows
| Measure | Figure | Source |
|---|---|---|
| Spaza shops in South Africa | ~150,000 | Accenture; Trade Intelligence. An estimate, not a census |
| Informal grocery sector value | R178bn – R207bn | Trade Intelligence, revised upward repeatedly 2023–2024 |
| The same sector, valued in 2019 | ~R40bn | Competition Commission, Grocery Retail Market Inquiry |
| Regular informal-channel shoppers | ~11.1 million | Trade Intelligence |
| Share of national food spend | 30–40% | Accenture; Trade Intelligence |
Those first and third rows disagree by a factor of four to five. That is not a rounding error — it is an order-of-magnitude dispute about a sector everyone agrees is enormous.
The reason is simple: South Africa has never conducted a retail census. Different studies count different things. Some include container shops and spazarettes; some do not. Some include hawkers; most do not.
When the 2024–25 registration drive ran, roughly 82,000 shops registered nationally — and analysts noted that this undercounts the real population substantially. The state itself does not know how many spaza shops exist.
Four structural weaknesses, and which one comes first
Every serious study converges on the same four. The Competition Commission's Grocery Retail Market Inquiry, Accenture, Trade Intelligence and the academic literature all arrive here independently.
No buying power. An individual shop buys at or near retail from a cash-and-carry. National chains buy the identical product materially cheaper. The Competition Commission identified this differential treatment by suppliers — particularly in the granting of rebates — as a central driver of chain dominance.
No working capital. Cash-based, no financial records, no credit history, therefore no bank finance. A stock-out is a cash-flow problem, not a demand problem.
No collective action. The Commission was blunt: locally-owned spaza shops largely fail to cooperate with one another, and so cannot access the bulk-buying prices that cooperation would unlock.
No records. Most owners do not know their margin per product, their fastest-moving line, or how much is out on the credit book.
The fourth is the precondition for the other three. You cannot negotiate volume without demonstrable volume. You cannot borrow without a transaction history. You cannot organise collective buying without knowing what everyone buys.
Source: Competition Commission, Grocery Retail Market Inquiry Final Report, December 2019.
Why nobody had solved it before 2024
Because until November 2024, no shop had a compelling reason to keep records.
Government has been trying to formalise township retail for over thirty years, and the pattern repeats with enough regularity to predict it.
| Intervention | Year | Instrument | Outcome |
|---|---|---|---|
| Businesses Act | 1991 | Deregulation | Weak municipal capacity; no register |
| Licensing of Businesses Bill | 2013 | Licensing | Stalled after criticism |
| NIBUS | 2014 | National support strategy | Built on a contested premise; limited measurable impact |
| Grocery Retail Market Inquiry | 2015–19 | Diagnosis and remedies | Excellent diagnosis, slow implementation |
| Gauteng TEDA | 2022 | Provincial legislation | Constrained by municipal capacity |
| Presidential directive | 2024 | Emergency regulation | Registration achieved; licensing did not follow |
| Spaza Shop Support Fund | 2025 | R500m blended finance | Slow disbursement |
Six attempts across three decades. The diagnosis is consistently good. Delivery is consistently the failure point.
The recurring constraint is not the legislation. It is municipal execution capacity — and in 2022, business owners in Diepsloot told the then-Premier of Gauteng exactly that: the law would change nothing if electricity, water, safety and refuse were not fixed first.
The R500 million fund, and where it went
The Spaza Shop Support Fund was announced in November 2024 and launched in April 2025, led by the Department of Small Business Development with the dtic, and implemented through SEDFA and the National Empowerment Fund.
It funds stock, refurbishment, training, compliance support — and, explicitly, point-of-sale devices, inventory management tools and digital payment systems.
| Measure | Figure |
|---|---|
| Fund size | R500 million |
| Businesses approved | 2,369 |
| Total approved value | ~R179.6 million |
| SEDFA | 1,316 applications · R79.6m |
| NEF | 1,053 enterprises · R99.9m |
To qualify, an owner must be a South African citizen or naturalised before 1994, operate in a township or rural area, be registered with the municipality, and hold a valid trading licence or permit.
That last condition is the whole problem. The fund is gated behind exactly the document 67,000 shops cannot obtain.
Sources: DSBD and dtic publications, 2025; spazashopfund.co.za.
What happened everywhere else
Between 2020 and 2022, African B2B platforms serving informal retailers — MarketForce, Wasoko, MaxAB, Sabi, Alerzo, Twiga — raised over $400 million on a single model: sell goods to the shops.
RejaReja, MarketForce's marketplace, reached 270,000 merchants across five countries. MarketForce raised $42.5 million, including a $40 million Series A at a valuation above $100 million. In April 2024 it shut down.
Its founder's own post-mortem named the causes: razor-thin FMCG margins, extreme price elasticity, and unit-level unprofitability. Copia collapsed. Zumi, Wabi and Capiter shut. Twiga cut over 850 staff. The survivors pivoted toward embedded finance, because financial services carry better margins than moving goods.
Five lessons that the survivors paid for:
- Distribution margins cannot carry technology-company burn
- Discount-driven acquisition buys volume, not retention
- Capital expenditure is the killer — trucks and warehouses are fixed costs
- Lending is a separate, hard business. Originate; never carry the risk
- Being asset-light is a decision that must be defended repeatedly
Sources: Tesh Mbaabu, shutdown statement, April 2024; TechCrunch and TechCabal reporting, 2023–24.
Why South Africa is different — and where it is harder
Three conditions exist here that did not exist in the Kenyan market when that category collapsed:
A legal forcing function. Kenya had no equivalent of the November 2024 directive. South African shops now have a legal reason to hold records.
A state fund that explicitly pays for the tools. Kenyan platforms had to convince shopkeepers to pay out of thin margins.
Pre-educated payments infrastructure. Flash, Kazang, Shop2Shop, Yoco and iKhokha have already normalised devices and digital rails in the informal channel.
And one condition that makes it harder. Formal retail is far stronger here. Shoprite's Usave eKasi, Boxer's trader ordering channel and Pick n Pay's township strategy are all pushing into the same neighbourhoods. The South African spaza owner is under more competitive pressure than a Kenyan duka ever was.
On ownership
Ownership of spaza shops shifted substantially toward foreign nationals from roughly 2005 onward. The research on why is unambiguous, and it is not about nationality.
Work by Charman, Piper, Liedeman and Petersen of the Sustainable Livelihoods Foundation, principally in Delft, Cape Town, attributes the outperformance to social networks delivering concrete commercial advantages: collective purchasing, shared distribution, pooled access to capital, and trusted low-cost labour. They framed the contrast as survivalist versus entrepreneurial business models.
Their later work found no clean link between nationality and price competitiveness, but did find a close correlation between competitiveness and the level of crime a shop experienced.
Two further facts belong alongside those. Foreign-owned spaza shops are estimated to pay in the region of R25 billion a year in rent to South African landlords. And the President stated in November 2024 that there was no evidence the food-safety problem was confined to foreign-owned shops.
The observed dynamic is about business model and social capital, not nationality. A South African-owned shop with access to collective buying and pooled capital would behave the same way. That gap is the one worth closing, and closing it is a construction case rather than a removal case.
Sources: Charman, Petersen & Piper (2012); Liedeman, Charman, Piper & Petersen, Econ3x3 (2013); Piper & Charman on xenophobia and price competition.
What is still missing
Nobody has a reliable, current, verified dataset of who trades where and what they sell. Not the state, not the manufacturers, not the banks.
Every intervention of the last thirty years has been designed in the dark. That is why they fail, and it is the gap this company exists to close.
This page was compiled in August 2026 from public sources. Figures, deadlines and programme details change — verify anything you intend to quote.