The Money You Can See

 

The last piece was about capital nobody could see. This one is about capital everybody can: the branded fridges, the free delivery trucks, the billions in corporate "empowerment" poured into South Africa's taverns. And why, after all of it, the industry has endured without ever really growing.

Second in a series of three. The first, The Money You Can't See, argued that immigrant traders beat local competitors on pooled, patient, invisible capital rather than on grit. This one takes the argument a step further.

---

Walk into a licensed tavern in Soweto or KwaMashu and the money is impossible to miss. The fridge behind the counter carries a brewer's colours. The board outside, with the tavern's own name on it, sits against a branded backdrop somebody else paid for. The chairs match. The glasses match. If the timing is right, a truck in the same livery pulls up and a driver wheels in the week's stock on credit, no cash changing hands. To a customer it looks like a well-run little business with a big sponsor. To anyone who reads value chains for a living, it looks like something else: a distribution point owned by one company and run, at thin margin, by another person entirely.

That gap is the subject of this piece. The first article argued that local entrepreneurs lost the spaza game because they could not reach the kind of capital immigrant traders reached: pooled, patient, and invisible from the street. The obvious objection writes itself. Fine, so give local business capital. What happens then?

South Africa has been running a version of that experiment for decades, in full view, and the tavern is where the answer shows up most clearly. Here is a sector that has had capital directed at it for years. Corporate capital, at real scale, from the biggest companies in the country. And it has not built the wealth its longevity should have produced. It has weathered almost everything thrown at it and grown remarkably little. That points to a conclusion worth sitting with: access to capital was only ever half the question. The other half, the half that decides whether money builds anything, is how that money is deployed, and on whose terms.

An industry that should have grown

Start with the size of it, because the numbers make the standstill harder to wave away.

South Africa's taverns turn over somewhere between R40 and R60 billion a year. They account for four-fifths or more of all alcohol sold in townships, and something close to 43% of the alcohol sold in the country as a whole. Counting the outlets precisely is oddly hard, because liquor licensing runs province by province and the national data is thin; the industry itself has long complained that nobody really knows how many licensed outlets there are. The National Liquor Traders Council speaks for more than 34,500 traders, but that is a membership roll, not a census, and the licensed total sits well above it once every province's taverns, pubs and bottle stores are added in. Behind the licensed trade lies a much larger informal one: researchers have put the number of shebeens, most of them unlicensed, at somewhere between 190,000 and 265,000. Together these outlets reach several million people a month and support well over a hundred thousand direct jobs. The spaza sector is larger by turnover, so taverns are not the biggest last-mile network in the country. But they are the second, and they are comfortably the largest last-mile distribution network in South African ownership. No locally owned retailer, courier or branch network sits closer to more South Africans, more often, than the tavern down the road.

By turnover the tavern sector sits second to spaza retail. What sets it apart is ownership: roughly 95% of taverns are Black South African-owned, and more than half are owned by women. This is the country's largest home-owned distribution footprint.

These are not fledgling businesses either. The typical tavern owner is in their mid-fifties. More than half of these outlets have traded for over a decade, many of them far longer, with roots that run straight back to the apartheid-era shebeen a grandmother ran out of a front room. By the ordinary logic of business, a sector this old, this large and this woven into daily life might be expected to have thrown off its own institutions by now. Locally owned wholesalers. Distribution companies. Buying groups with real weight at the negotiating table. Perhaps, at the far end, a brand or two to call its own.

Almost none of that has appeared. The sector has held its ground rather than climbed. Decades of trading, and the tavern of 2026 sits on much the same rung it occupied in the late 1990s: a single outlet, selling the manufacturers' products, at prices those manufacturers recommend, on terms they set. That is endurance without accumulation. And it would be a serious mistake to read it as a failing of the people behind the counters, who have kept these businesses alive through conditions that closed far sturdier enterprises. The question is not why tavern owners fell short. It is why so much invested capital left them exactly where it found them.

Because the one thing that cannot be said is that nobody put money in.

The most heavily backed small business in the township

If anything, the tavern is the most invested-in micro-enterprise in the country. No other township sector has drawn anything like the capital and corporate attention that has flowed into this one, and it has come not from one company but from the liquor industry as a whole.

South African Breweries, the dominant player and the one that pioneered the model, is the clearest case. It spent R54 million putting more than 16,000 tavern owners through its Mahlasedi Taverner Development Programme between 2004 and 2009, then followed it with a Customer Business Development course reaching around 4,200 more. Its Responsible Trader Programme has accredited over 32,000 traders. It hands out branded fridges, boards, chairs and glassware to outlets across the country. Its owner-driver scheme, running since the late 1980s, has put hundreds of Black South Africans behind the wheel of trucks they own on paper. And it extends credit to almost every licensed outlet in its network, so stock arrives today and is paid for later.

It is not alone. Heineken Beverages, formed in 2023 out of the merger of Heineken South Africa, Distell and Namibia Breweries, runs a Tavern Transformation Programme that has rebuilt more than 240 taverns across eight provinces, fitting kitchens, toilets, CCTV and digital screens, with McCain, Primedia and Vodacom as partners and Amstel Lager as brand sponsor. Its Responsible Trader training has passed some 3,000 owners, and its reward programme flies compliant traders to Cape Town and hands out the occasional bakkie. Distell, before the merger, built Click2Collect to route consumer orders through more than 22,000 taverns during the pandemic. Spirits houses like Diageo and Pernod Ricard run their own trade programmes, and specialist firms such as Touchsides have spent years wiring up tavern tills to harvest and sell the sales data. Then there are the ownership schemes on top of all this. SAB Zenzele matured in 2021 as the largest broad-based empowerment deal in the history of South African consumer goods, worth R9.7 billion at the end; a retailer who put in R100 in 2010 walked away with a pre-tax R77,518. Its successor, Zenzele Kabili, holds R5.4 billion in the parent brewer's shares and lists on the JSE.

And all of that, the named programmes and the empowerment schemes, is only the visible tip of the spend. It sits on top of a far larger, continuous outlay that seldom makes the press releases. Measured advertising alone, the television and radio and billboards, reached about R1.7 billion a year over a decade ago, with the big brewers and Distell accounting for roughly 85% of it, and it has grown since. Even that figure stops at the media line. It leaves out the money spent inside the outlets themselves: the fridges and coolers, the branded shopfronts, counters and furniture, the umbrellas and glassware, the point-of-sale material, the trade-marketing teams and merchandisers, the in-tavern promotions and sampling. A single branded cooler costs several thousand rand; place and service them across a six-figure base of outlets, year after year, and the fittings and below-the-line spend alone run comfortably into the hundreds of millions annually, before a cent of measured advertising is counted. The tavern is not lightly sponsored. It is one of the most intensively marketed retail environments in the country.

Layered together, this is an extraordinary amount of capital, skills and infrastructure aimed at one small-business sector over two decades. It is worth pausing on how unusual that is. Branding and equipment are not, of course, unique to liquor. Walk into any spaza and you will find a soft-drink company's fridge, a snack brand's shelf strip, a mobile operator's airtime board. Consumer-goods manufacturers and network operators put plenty of kit and livery into township retail. But that support is mostly marketing: paint the wall, chill the drinks, move the product. It rarely comes with the depth of the liquor model, where credit, exclusive equipment, training, distribution and even shareholding all bind the retailer to the supplier at once. The tavern did not merely get branded. It got wrapped.

And still it did not grow. Which is the puzzle. If capital by itself were the missing ingredient, this is close to the opposite of what should have happened. So the interesting question is not how much money went in. It is what shape that money took on the way.

The chain governor

To see the shape, you have to see how the liquor value chain is built, because it is built to concentrate power at the top.

Economists have a clean name for an arrangement like this: a captive value chain. One dominant firm sits at its head, sets the rules, and decides who is allowed in. A Wits University study of the brewer-shebeen relationship describes the leading brewer in exactly those terms, as the "chain governor" that concentrates power and makes the decisions that matter about who gets included. The picture it paints is stark. The brewer contracts the farmers who grow the barley. It brews at industrial scale, on the order of billions of litres a year, and runs the stock out through dozens of depots and a fleet of trucks. And it lands that product in the taverns and shebeens that retail the large majority of everything it makes.

Profit concentrates where brand power, scale and capital intensity live: at the manufacturer and big-retailer end. Township outlets, taverns among them, sit at the bottom, distributing goods made elsewhere and keeping the thinnest slice.

Every link in that chain is tuned for one outcome, which is the manufacturer's volume. The tavern owner is not really a partner in the arrangement, whatever the branded board out front suggests. The tavern is the captive endpoint where the product finally meets a customer. And a captive endpoint is a strange place to aim "empowerment" capital, because in a captive chain the lead firm has little reason to fund a retailer's independence and every reason to fund its reliability. It gives enough to keep the outlet trading and stops short of anything that would help it leave. That is not a morality tale about wicked brewers. These are rational companies doing what the structure rewards, and they would be failing their own shareholders if they did otherwise. The trouble lies in what the structure rewards, and in the fact that no countervailing structure was ever built on the tavern's side.

The fridge was never his

Look closely at each piece of "support" and the same design shows through. Every one of them, whatever else it does, tightens the knot.

Take the fridge, the most visible gift of all. It is not a gift. It stays the manufacturer's property, on permanent loan, and it comes with rules. Only that company's beer may go in it. A branded fridge must show that brand on the top shelf. It must stand where customers can see it. Stock a rival's product and the fridge can be taken back. The Wits researchers recorded a Soweto owner who moved to a competitor's range to chase a younger, premium-drinking crowd and watched the brewer come and remove its fridge. There are, as the study put it plainly, consequences for divorcing your supplier.

Then the credit, which binds harder than any contract. Nearly every licensed outlet orders on account, on seven, fourteen or thirty-day terms. It reads like a favour and works like a leash. One shebeen operator in the study described the brewer as a bank: all the shebeens order their beer on credit. One week the delivery simply did not come, a fault somewhere up the supplier's own chain, and he sat idle for days rather than walk to a nearby wholesaler and buy stock to trade. The reason was almost mundane. He had already paid a large sum into the brewer's account and had not budgeted to buy beer twice. The credit had folded him so completely into one company's logistics that sourcing beer anywhere else had stopped being a thought he could reach for.

The owner-driver scheme runs on the same logic dressed as liberation. A driver buys a branded truck on finance the company arranges, earns a genuinely respectable income, and is celebrated as a business owner holding the keys to his own future. Read the terms, though, and he services one company only, is barred from carrying anyone else's product, and has his routes and his day set for him. The scheme trimmed the brewer's distribution costs sharply while keeping tighter control than it would have held over its own staff, because the driver carries the capital risk and can be dropped if he steps out of line. It is an elegant machine, and it is elegant chiefly for the company that designed it.

Even the training carries the pattern, and this is the quietest part of it. The courses are real, the certificates accredited, the record-keeping skills genuinely useful. But notice what is never on the syllabus. Nowhere does a tavern owner learn to negotiate across several suppliers at once. Nowhere does he learn to pool his buying with the four taverns on the same street to force a wholesale price. Nowhere does he learn to diversify out of the single product category the manufacturer happens to control. The training turns out a more capable retailer for the brewer. It was never designed to turn out an independent business, and it would be strange to expect the brewer to fund the one course that might teach a tavern to need it less.

The architecture, not the amount

Pull back from the detail and the principle fits in a sentence. What decides the outcome is not how much capital arrives. It is the architecture of it: the terms it carries, and who keeps control once it lands.

The first article introduced that phrase, capital architecture, almost in passing. This is where it earns its place. The capital described in that piece was of an entirely different kind from the capital described here. It was pooled through rotating community credit and family networks; it was patient; and, decisively, it left control with the person who received it. Nothing in it dictated which supplier to use, which brand to stock, or what happened to your equipment if you switched. It carried heavy social obligations, but those obligations pointed outward, into a widening set of options: the freedom to experiment, to diversify, to fail cheaply and begin again with a different supplier.

Manufacturer capital is engineered the other way. Every element of it, the fridge, the credit line, the sales rep's attention, the shares, deepens reliance on one company's products, logistics and systems. The fridge is conditional. The credit is tied to the brand. The training teaches the supplier's way of working. Even the shareholding hands the owner a sliver of the very corporation whose position his dependence secures. It is a whole system of managed support, and its subtlety is that from the inside it feels like being looked after, right up until the moment you try to leave.

One design ends in the freedom to walk away. The other ends in a business that cannot be left without surrendering the assets that made it viable. That single difference does most of the work.

It is worth being clear that this does not contradict the first article; it extends it. That piece said local entrepreneurs lost out because they could not reach the right kind of capital. This one says that even when local businesses reach capital in abundance, its kind and its terms decide what it builds. Access is the first gate. Architecture is the second. A sector can clear the first and still be stopped cold at the second, and the tavern is what being stopped at the second gate looks like.

The clearest proof is what the tavern never became. The move that would have changed its fortunes is what economists call backward integration: pushing upstream, out of pure retail and into wholesale, distribution, or eventually the making and branding of product, so that value is captured at more than one point in the chain. That move has been made elsewhere in South African township retail, by operators who used capital they controlled to buy in bulk, run their own distribution and open at scale. The tavern sector, despite decades of trading and a fortune in corporate backing, has not made it at all. It sits today exactly where it sat at the start: a retailer of other companies' brands, at the bottom of a chain those companies govern. Not for want of capital. For want of capital it could point wherever it chose.

Endurance is not the same as growth

There is a temptation to look at a sector that has lasted this long and call it a success simply because it has lasted. That temptation is the trap.

The tavern trade has passed every survival test the country could set it. It survived criminalisation under apartheid, when the shebeen was an act of quiet defiance run out of sight of the police. It survived legalisation and the wave of corporate attention that followed. It survived the COVID prohibition, when the government banned alcohol sales outright and weekly takings collapsed, and it came back anyway. It survives a steady weather of new regulation, licensing battles and public hostility. When one brewer unveiled its tavern-transformation drive, some of the loudest replies were not gratitude but grief and anger, people recalling a parent killed in a tavern, or warning that this was simply big alcohol dressing expansion up as upliftment. Those are serious objections, and they deserve a serious answer. The industry absorbs all of it and keeps trading.

But endurance and growth are not the same thing, and the distance between them is where the real loss sits. This sector persists without accumulating. It employs people, pays rent, feeds families, and at the end of every year most of the value it has moved has flowed straight through it and out. Research on township economies keeps finding the same thing: only about a quarter of the money generated in these communities stays in them, with the rest draining up the chain to manufacturers, wholesalers and head offices elsewhere. The tavern is one of the widest pipes that drainage runs through. It distributes wealth made elsewhere and keeps a sliver for the effort. That is what capital deployed on someone else's terms finally buys: permanence without progress. A sector that cannot easily die and cannot easily grow.

So what is a tavern, actually?

This is where the whole thing is worth turning over, because the debate may have been fixed on the wrong object all along.

The tavern is usually argued about as a place to drink, with the argument then running to whether that is good or bad for the community around it. Set the liquor aside for a moment and look at what remains. Tens of thousands of owned premises. Embedded in the last mile, in exactly the communities that formal retail, banking and logistics find hardest and most expensive to reach. Locally trusted. Staffed. Open long hours. Holding, between them, the largest South African-owned distribution footprint in the country. Read that list back. Not one line of it describes a bar. Every line of it describes a distribution network. In that respect a tavern is no different from a spaza, or from the bottle store in the shopping centre: at bottom, it is a point through which goods and services reach people.

The formal economy has always understood this about liquor, even where the township debate has not. In any shopping complex, the bottle store sits a few doors from the supermarket. The restaurant that pours wine opens where the shops and the foot traffic already are. Liquor retail has never stood apart as its own peculiar category; it clusters with groceries, with services, with everything, because it was always one more node in a wider web of distribution. The tavern is that same node, sitting in the township rather than the mall. One product has been allowed to define the whole of it, and the definition has been treated as fixed.

Loosen it, and the questions change completely. The interesting one stops being how to make taverns better bars and becomes something else: what could the country's largest home-owned last-mile network carry, if the capital behind it were deployed to build the network rather than the dependency? Groceries on the way home. Airtime, data, electricity, a place to send or collect money. A parcel to pick up. Connectivity for a street that has none. And beneath all of it, the one move the sector has never made: pooling its buying power so that, on the single product it does sell, it stops being a price-taker and starts negotiating like the multi-billion-rand channel it actually is.

There is even a way out of the harm trap in this. A tavern defined only by drink is permanently one ban, one backlash, one moral panic away from ruin, and its critics are not wrong to worry about what it can do to a neighbourhood. But a tavern that is also where the community collects its parcels, tops up its electricity and gets online is something else. It is something a community has reason to defend rather than resent, and something whose owner is no longer staking an entire livelihood on how much the neighbours drink.

Can a tavern be more than what it is today? A proposal built on exactly that question was in fact put to the liquor industry and its manufacturers in 2020, at the worst possible moment for the sector, when taverns were shut by the trading ban and desperate for a future. It argued that taverns could shift from being social hubs to being economic ones, and it asked the manufacturers to get behind that shift and fund it. For a while there was real interest. Then the ban lifted. What happened next, and what it revealed about who these networks are really built to serve, is the subject of the final article in this series.

---

The back story

I worked for the South African Breweries for a few years, and after I left I kept working the same trade, providing trade-marketing services into the liquor industry. Later I worked more closely with the tavern owners themselves, through their associations, close enough to hear how the arrangement felt from their side of the counter.

Beyond the time at SAB and in trade marketing, I was, in 2020, commissioned to present a report on the tavern industry's potential to shift from social hubs to economic hubs, at the conference that formed the National Liquor Traders Council. The findings of that report and my vision of how taverns could play a role in transporting the township economy, are the third and final article in the series.

A caveat worth keeping in view: everything here describes the industry at the level of patterns, not individuals. Many tavern owners run excellent, disciplined, much-loved businesses, and nothing above is a judgement on any one of them. Reframing the tavern as a distribution network is also not a way of waving off the real harm alcohol does in townships. It is, if anything, the opposite: an argument for a model that no longer depends on that harm to survive.

Research. The deep research behind the value-chain and empowerment-scheme material was done with Kimi Deep Research and NotebookLM. Current developments, including the Heineken Beverages tavern-programme figures and the disputed 2026 Zenzele Kabili shareholder meeting, were checked with Claude (Opus 4.8 and Sonnet 5) using live web search. Where a claim is contested, such as the allegation that some Zenzele Kabili shareholders went unpaid for years, which SAB denies, I have said so rather than pick a side. Some figures from the older industry material were cross-checked and updated where they had dated.

Writing. Drafted and edited with Claude (Opus 4.8 and Sonnet 5).

Images and video. The data charts were built in matplotlib (Python). The hero image was generated with Nano Banana Pro (Google Gemini 3 Pro Image) on kie.ai, run through an n8n workflow.