🍞 The Social Game

📈 Stocks
Nigerian equities slipped below 250,000 and closed Wednesday at 247,673.65, down about 1% on the week. Trading was thin, with almost every stock unchanged on the day, so investors look to be waiting rather than selling hard.

💱 Currency
The naira held steady at around ₦1,333 to the dollar. The Central Bank of Nigeria (CBN) said gross reserves reached $55.07 billion, the highest in years.

🛢️ Commodities
Brent climbed back above $100 a barrel this week after a container ship sank in the Strait of Hormuz and Houthi forces kept attacking Saudi Arabia. Prices eased to around $104 on Thursday morning on profit-taking.

On 26 September, Opeyemi Adeniran, a Nigerian frontend engineer living in Berlin, posted a browser game called Lagos Run (now Danfo Run). You drive a yellow danfo from Ojuelegba to Yaba, pick up passengers, dodge okadas and honk your way through traffic while a fictional station, 97.3 Ojuelegba FM, plays Afrobeats. He built it in four to five weeks after work, using AI tools including Claude for both the code and the graphics. There is no login, no download and no paywall. Within days, the game had drawn tens of thousands of players. By early October, reported figures put it above 140,000, across 145 countries.

Five days after Danfo Run went live, on Independence Day, Shalom Rayhamen, a UK-based Nigerian software engineer, launched Lagos Life. She had seen Lagos Run on X and built the first playable version of her own game in about eight hours on a Wednesday night. Lagos Life is The Sims set in Lagos. You choose to start as "Nepo" (wealth and connections) or "Lapo" (hustle), get a job, pay rent in virtual naira, watch your energy bills, and bump into thousands of other players at virtual versions of real-life nightclubs like Quilox, Amala Shitta or by the lagoon.

The game grew to 1 million players in three days, 5 million by 9 October and 10.15 million by 11 October. By day nine, Rayhamen had incorporated Vatar Inc. in Delaware and closed a $500,000 angel round from three founders: Olugbenga Agboola of Flutterwave, Abdulhamid Hassan of Mono and Nathan Nwachuku of Terra Industries. The round reportedly values Vatar at $10 million.

A week on one of the game's virtual airport billboards now costs ₦500,000 ($375). At launch, a billboard cost ₦5,000. At the time of funding, the company said it had $170,000 in booked revenue, primarily from ads, but now plans to monetise through more typical social gaming strategies.  

There is only one question on our mind…why now? As far as I can tell, there are three reasons.

A game on a phone is mostly maths for the graphics chip, or GPU. Sixty times a second, the chip works out where every shape sits on screen and what colour each pixel should be. For most of the internet's life, a webpage could not talk to that chip directly. Serious games had to be downloaded as apps or played on gaming consoles. 

That began to change in 2011 with WebGL. It lets the code inside a webpage send small programs, called shaders, straight to the GPU. WebGL 2 followed in 2017 and is now supported almost everywhere. It is why a Sims-style city can run in a phone's browser in the first place. 

The next step is WebGPU, a newer standard modelled on the graphics systems that native games use: Metal on Apple devices, Vulkan on Android, Direct3D on Windows. Its big addition is compute shaders. These let the GPU do general work, such as in-game physics, crowd movement or running a small AI model, not just drawing pictures.
Fun fact: before the AI wave, Nvidia's primary business line was creating these GPUs for gaming companies. 

Chrome shipped WebGPU in 2023. The last major holdout fell in September 2025, when Apple added it to Safari on iPhone, iPad and Mac. It now runs by default in every major browser, and engines such as Three.js, Babylon.js and Unity can target it. Developers still keep a WebGL fallback, because older Android phones depend on what their graphics chip supports.

Three other pieces finish the job:

These are critical in a real-time social game that is essentially an always-on live event.

Lagos Life had to rebuild its plumbing in the middle of going viral. Rayhamen says she started on Cloudflare's database service and found it could not cope. She moved the database to DigitalOcean, then moved the whole app to Amazon Web Services with autoscaling (servers that add capacity automatically as players arrive). From our data, at its busiest, the game has had about 180,000 players online at once. For a Nigerian player, the browser removes three barriers at once: the app store, phone storage and a card payment. The entire distribution strategy is a link. Click and you can start playing immediately. 

The second shift is in who can build. Making a game used to need a team: programmers, artists, sound designers and a producer, working for a year or two. Adeniran built Danfo Run alone in about a month. Rayhamen had Lagos Life playable after one night, and she now uses AI agents to catch bugs and sort feature requests from millions of players.

The big platforms are racing to make this normal:

Cristóbal Valenzuela, chief executive of the AI video company Runway, describes the shift as a move from rendering pixels to generating them.

A rendered world is fixed: every corner has to be built in advance. That is why rich interactive worlds have usually cost either a fortune, like Grand Theft Auto, or looked basic, like early Roblox. A generated world can be extended relatively cheaply, first by its creator and eventually by players.

Music production has been through a version of this. Once digital audio workstations put a recording studio on a laptop, the number of songs being created exploded and, more importantly, access to music creation expanded. Afrobeats as we know it today is a direct result of this democratisation of music creation. We are beginning to see the same in gaming and films. 

When anyone can build the game, the game stops being scarce. What stays scarce is creativity around game dynamics and distribution. Anyone who has watched a 12-year-old build something in Roblox has already seen this happening.

The players were ready before the tools were.

In Deloitte's 2021 Digital Media Trends survey of US consumers, video games were Generation Z's favourite form of entertainment, chosen by 26%. Watching TV shows or films at home came fifth, at 10%. For Boomers, it was the reverse: 39% picked TV and films, and only 3% picked games. Millennials sat in between, at 18% for TV and films and 16% for games.

More recent data shows the drift continuing. Research firm MIDG adds up average daily viewing across the five biggest platforms in each category, among their users. The totals are not one person's day, but the direction is clear:

Gen Z is a social media generation; they expect entertainment experiences that are shared with their friends. Gen Alpha is an interactive one: it expects to talk back to what it watches. A shared virtual city, where you can throw a party, play your song or advertise your business, fits that habit better than a two-hour film does.

There is a Nigerian twist, and it explains Lagos Life's business model. In 2023, only 1% of Nigerians paid for a music subscription, according to IFPI and World Bank data. In the US the figure was 42%, and in Sweden, 50%. Nigerians will spend hours on entertainment, but very few pay for it directly. A free game that charges brands rather than players fits that market.

Lagos Life makes most of its money by selling the city's attention to brands. According to the game's ad page, checked on 9 October:

Musicians have paid ₦50,000 to have their songs played in the game's clubs. Vatar also sells branded danfos, storefronts and company headquarters inside the game. Players can top up their in-game cash with real naira. Rayhamen says these bundles started at ₦500 and now go up to ₦20,000.

On revenue, Rayhamen says the game made $46,900 in its first four days and more than $100,000 by 7 October. Vatar now says it has earned and contracted more than $170,000, a figure that mixes cash received with deals signed. Against estimated hosting costs of roughly $6,000 a month, that is a strong start. It is also very small, next to the companies whose playbooks it is borrowing. Which is where the opportunity is. 

The Chinese media conglomerate, Tencent, owner of WeChat and QQ, began as a messaging app. ICQ. Its breakthrough came from selling virtual goods, not ads. In 2002, it launched  QQ Coins, a virtual currency priced at one yuan per coin. Later on, it launched QQ Show. Based on South Korea's Cyworld, it let users dress their avatars. For many teenagers, clothes for their digital selves stood in for clothes for their real ones.

By the early 2010s, value-added services paid for in QQ Coins made up more than 80% of Tencent's revenue. The coins became so widely accepted that China's central bank tried to rein them in when shops started taking them for real goods!! 

In 2025, Tencent's domestic games alone earned RMB164.2 billion (about $23 billion) and its international gaming revenue passed $10 billion for the first time. QQ, although not the main revenue driver of Tencent, still has 508 million mobile users a month.

Advertising, by contrast, made up about  20%  Tencent's revenue in the third quarter of 2025. 

Minecraft charges once, up front, then sells extras.

Microsoft paid $2.5 billion for Mojang in 2014, when the game had sold about 54 million copies. By April 2025 it had sold more than 350 million, making it the best-selling game ever: more than six times the base Microsoft bought.

On top of the game sits the Marketplace, where outside creators sell maps, skins and mini-games. The marketplace alone generated more than $500 million by 2022. Then came subscriptions for private servers, licences for schools, and a 2025 film that took close to $1 billion at the box office. The world itself became IP that could be extended beyond the game. 

Roblox is arguably the most popular gaming platform in the West. It is free to play. Players buy Robux, the platform's currency, and spend them inside games that other people build. Roblox keeps the money from Robux sales and pays creators through its Developer Exchange, which converts the Robux they earn into real money. In 2025, players bought $6.8 billion of Robux, and the company paid creators $1.5 billion, about 22%. Its top 1,000 creators earned an average of $1.3 million each, making being a Roblox creator a highly lucrative role. 

Advertising is a small, secondary line.

This is the opportunity ahead for Nigeria’s new social gaming upstarts: if they can turn attention into habit and habit into a user-controlled economy, there is a lot of money to be made. 

Most social games that work long term have the same five traits:

Those five make a good game. A platform needs one more: other people must be able to build on it and get paid. Roblox is a platform because a teenager in Indonesia can build a game there and withdraw real money. Minecraft became one when it opened the Marketplace. Highrise followed in 2024 with Highrise Studio, which pays creators based on item sales and engagement; its player-run marketplace has passed $250 million in cumulative sales.

Plenty of social worlds hit numbers like Lagos Life's and still died.

Rec Room is the clearest recent example. Founded in 2016 by Nick Fajt and Cameron Brown, it attracted more than 150 million players at its peak and was valued at $3.5 billion in December 2021. It shut on 1 June this year. Revenue never caught up with the growing costs of serving its 150 million players. The company said a shift in the VR market, plus wider headwinds in gaming, had made profit too hard to reach. This is a common fate for viral apps. Famously Snap CEO speaks about the curse of having an extremely popular app which costs a lot to run, without the right infrastructure to reduce server bills or monetize adequately. This was one of the reasons why Youtube sold to Google (owner of Google Cloud); Twitch to Amazon (AWS) and Instagram to Meta (Facebook’s existing infrastructure)

Safety is also a major challenge. Roblox faces more than 140 lawsuits alleging that adults used the platform to groom and exploit children, and several US states have sued it. Its response, mandatory facial age checks for chat from January, has a real cost. Daily users fell from 152 million in the third quarter of 2025 to 123 million in the second quarter of 2026, and the company cut its 2026 bookings forecast by about $900 million. Safety is expensive to build and more expensive to skip.

Lagos Life hit the same wall in its first fortnight; A website advertising sex work bought billboard and lagoon space. The ads came down within a day of being reported. An 18+ rule. The team has had to block accounts impersonating it and warn players about a $LAGOSLIFE crypto token it did not create. And of course players cheat; It wiped balances after some players found a way to create what Rayhamen called "trillions" in virtual cash.

Every one of these problems grows as the game does.

Danfo Run and Lagos Life prove two things: the cost of building a social world has collapsed, and Nigerians will turn up in their millions for one that looks like home. The lasting value sits with whoever owns the venue that others build in and earn from, not with whoever throws the first viral party.

If this is useful, don't gatekeep. Forward this to one person who needs it. That's how we grow 🍞

Dangote Petroleum Refinery made a profit after tax of $4.73 billion in the first nine months of 2026, against a loss of $117.2 million a year earlier. Revenue more than doubled to $22.48 billion. About $2.91 billion of that profit came in the third quarter alone, a margin of roughly 34%. The plant now runs at its full 700,000 barrels a day, and net debt of $1.73 billion has turned into net cash of $650.8 million.

The results land in the middle of the share sale. At ₦525 a share, the refinery is valued at about ₦65.2 trillion ($49 billion), or around eight times this year’s profit so far. The offer has been extended by two weeks.

Watch: the IPO closes on 27 October, with listing on the Nigerian Exchange expected in November.

👉 Read the full story

OPay has filed to list on the New York Stock Exchange, and its filing shows revenue more than doubled to $467.1 million in the first half of 2026. Net income jumped from $21.7 million to $90.9 million. The surprise is where the money comes from: lending brought in $236.1 million, about half of revenue, against $130.6 million from payments.

New loans tripled to $927.6 million, but growth has a cost. OPay set aside $118.4 million for bad loans, about a quarter of its revenue. Nigeria accounts for nearly 90% of its business, and the app has 50.1 million monthly users.

Watch: pricing is expected in mid-November, followed by a secondary listing on the Nigerian Exchange.

👉 Read the full story.

🇧🇼🇳🇬 Botswana becomes Southern Africa’s gateway to the Dangote refinery IPO
Botswana will host trading for the Southern African slice of the Dangote refinery IPO. Access Bank Botswana, the local arm of the Nigerian lender, will issue depositary receipts, which let investors own the shares without buying them on the Nigerian Exchange. It expects to raise $500 million this way. The region’s 14 stock exchanges agreed to route all trading through one market “to reduce any complications”, said Collen Tapfumaneyi, who chairs the SADC Stock Exchanges Committee. The IPO is selling 4.1 billion shares at ₦525 each to raise about $1.6 billion. That could reach about $2.1 billion if extra shares are sold to meet demand. Speaking in Gaborone, Aliko Dangote said “we have more than reached our target”, with demand at three to four times the amount sought. The SEC has given the offer two more weeks, to 27 October. Watch: how many shares each market gets once the offer closes.

🇬🇳 Rio Tinto’s Simandou mine hits a 10 million-tonne pace
Rio Tinto says its SimFer mine at Simandou in Guinea reached a production pace of about 10 million tonnes of iron ore a year in Q3. The mine produced 2.5 million tonnes in the quarter, up 273% on Q2, and shipped 2.2 million tonnes. That pace is a sixth of the mine’s full 60 million-tonne capacity, which Rio aims to reach in the second half of 2028. Rio owns 53% of SimFer, and Chinese state miner Chalco owns the rest. Rio’s share of the building cost is $6.2 billion. The mine is about 84% complete and the port 92%, and the first ore is due to run through the permanent crusher and port this quarter. About 75% of the 17,430-strong workforce is Guinean. Rio expects to sell 5 to 10 million tonnes from Simandou this year. Watch: whether the crusher and port start working this quarter as planned.

🇧🇼 Botswana set to raise its De Beers stake to about 25%
Botswana is close to a deal that would lift its stake in De Beers from 15% to about 25%, according to people familiar with the talks. The deal would be with the Global Diamond Consortium, which also includes Namibia and Angola. It is led by former De Beers chief executive Gareth Penny and became the preferred buyer for Anglo American’s stake in July. Nothing is final yet. The three governments together would own less than 50%, so De Beers would stay privately controlled, which some consortium backers require. That falls short of President Duma Boko’s earlier push for majority control. Anglo has discussed selling for about $1 billion. It has written down De Beers’ value three times in three years, to $2.3 billion. Diamond prices have slumped as Chinese luxury spending collapsed and lab-grown stones took market share. Watch: a signed sale agreement, which would end a process that has run for more than two years.