All the answers {{ FIRSTNAME | there }}? Jip, students in Asia are apparently now renting smart glasses to cheat on tests. We want to go tsk-tsk, but if we were still in school, we’d prolly do the same thing. 😇

Right now, though, we’re getting ready for the long weekend. So, as a Happy Easter surprise to you, it’s another awesome founder story... with some extra goodies for you at the bottom. Enjoy! 🐣

While the rest of SA's fibre industry chased leafy suburbs, Alan Knott-Craig bet everything on a R5-a-day model for the 13 million homes they left behind…

Alan Knott-Craig didn't need to start another business. By 2022, he'd already built and sold HeroTel, SA's largest rural broadband provider. He had money, a good track record – you could have forgiven him for just retiring to the winelands to write more awesome books.

But then he wouldn’t have been Alan.

Instead, he decided to do the thing that no one in the country (as far as he can tell, the world) had figured out how to do successfully yet: Put fibre internet into every home in every township in South Africa.

It all started one day in 2010…

A two-year-old, an iPad and one unshakable question

Alan has been building telecoms businesses since 2003 (Cellfind, iBurst, Mxit, HeroTel), with some hits and some spectacular misses. But none with an origin story quite like this.

See, one afternoon in 2010, Alan was watching his two-year-old daughter sitting on the floor, swiping through an iPad, completely at home on the internet. And it hit him: Just down the road, there were families whose children didn't have connectivity at all. His daughter would grow up with a tremendous advantage, and millions of SA kids wouldn't, simply because of where they were born.

He imagined her at 25, asking him what he did with his life. And "I made a lot of money and built cool businesses" wasn't the only answer he wanted to give.

"Look, it's no secret; I want to make lots of money and do well for my family," Alan says. "But I played out the scene in my head where South Africa stays this divided, with such a huge part of the country unable to catch up because of a lack of infrastructure. And I just thought: If I want to stay in this country, if I want my children to be able to live here one day, something has to be done to level the playing field."

He calls it a selfish reason. And maybe it's a small thing; people need roads, hospitals and jobs, after all. But internet access was the one lever he knew how to pull, so it kicked off a business 10+ years in the making.

The winding road to fibre

We didn’t know it, but that newfound purpose shaped Alan’s every business decision from 2010 onwards. But it took over a decade of trial and error to arrive at the winner.

The short version: He bought Mxit, got the strategy wrong trying to compete with WhatsApp, got kicked out, lost a fortune, and spent eight years rebuilding from a rented cottage on a farm. He launched Project Isizwe (free public Wi-Fi in Tshwane), then built HeroTel into SA's largest rural wireless broadband provider over eight years. (For the full rollercoaster, watch our podcast conversation with Alan.)

But through all of it, one technical truth kept asserting itself: Wireless can't handle density. In a township like Alexandra, where over 150’000 homes sit on top of each other, you can’t have millions of people streaming a soccer match at 100 Mbps over the air at the same time. There's not enough space in the spectrum.

The only answer was fibre. So he sold HeroTel. And 5 weeks later, started Fibertime.

R5 a day, no contract, no nonsense

Fibre wasn’t new in SA. The conventional industry had spent 12 years connecting roughly 4 million homes, almost all of them in leafy suburbs, almost all on postpaid contracts. Meanwhile, 13 million homes in townships had nothing.

Alan's thesis was simple and borrowed from the oldest trick in the SA consumer playbook: Make it prepaid. Real prepaid; not R300 at the start of the month, but pay-as-you-go, R5 for 24 hours of uncapped, unthrottled 100 Mbps fibre. The same insight that transformed the mobile industry (from 3 million phones to 70 million once prepaid arrived) could transform fibre.

But the model required a completely different approach to everything. Telecoms is not a cheap game. And Fibertime couldn’t wait for customers to sign up; if this would work, they’d have to go into a township, plant poles, run aerial fibre and connect every single home with a free Nokia router and a battery backup (for load shedding — "otherwise, what's the point?"). 

And it wasn’t going to be with a proven take-up rate. Residents were going to buy a voucher at a spaza shop or through a banking app, punch in the code, and they're online.

So you can imagine a key question was: Who the heck was going to invest in this gamble?

"Telecoms is all about money, don't let anyone tell you otherwise," Alan says. "And the market we needed to go after has entirely different earning and spending habits. You can't think like a normal fibre provider that invades an area and aims for thousands. With this, you have to go in with a mindset of capturing millions."

Kayamandi and the funding wall

The first installation in Kayamandi, outside Stellenbosch, went live in late 2022 — funded largely out of Alan's own pocket from the HeroTel sale, plus some cash from his HeroTel investors. No one else would touch it.

Eventually, he got Rand Merchant Bank on board, partly because Fibertime's township model slotted perfectly into their internal ESG targets. Finnish development financier Finnfund came in with an initial €2 million (later doubling down with another €2 million). Nokia signed on as the infrastructure partner. And the first township proved the model.

The unit economics make sense precisely because townships are so dense: The cost to connect every home in a one-hectare block is roughly the same whether it's a suburb or a township — but there are far more homes (and therefore far more potential revenue) in the township. At R5 a day, a customer spends about R150 a month. If they consumed the same data on mobile, Alan estimates they'd pay over R1’600.

380’000 homes and counting

Today, Fibertime is live in 49 townships across 13 cities and towns in 7 provinces, with around 380’000 connected households — not "homes passed," which Alan points out is a metric the industry loves to hide behind, but actual connected homes with active routers.

They're connecting 1’200 households a day, and that number is growing by 20–40% month-on-month. To put that in perspective, Alan believes that once you cut through the industry's smoke and mirrors, Fibertime is probably close to being SA's third-largest internet provider by connected homes.

The Nokia partnership has been key to scaling. In October 2025, the two companies announced the rollout of an additional 400’000 homes, as part of Fibertime's larger goal of reaching 2 million connected households by 2028 (a target requiring around R10 billion in total investment) and then 5 million by 2030 (at R25bn).

And it's not just about routers and revenue. Because Fibertime physically enters every home during installation, they've built an unusual dataset: detailed household mapping, aerial imaging and bandwidth usage patterns across 265’000+ homes. Their insights team has started using this data to challenge official population figures, suggesting SA may be significantly more populous than census data reflects.

Respect is the real infrastructure

One thing Alan is adamant about: You don't build in townships without earning the right to be there.

"If you're not part of the community, there's an immune system that will expel you — violently, if they think you're just here to take advantage," he says. Fibertime hires local labour, works through church and school leaders, and has women from the community escort site visits. "They know everyone and can spot when something is about to go wrong, and they intervene before it does."

It's working. Despite operating in some of SA's most complex environments, Alan says vandalism and theft are virtually non-existent. When people realise their kids can study online, stream entertainment and stay safely indoors, the network becomes something the community protects, not targets.

What comes next

There's been talk of a JSE listing, potentially by 2027, though Alan frames it carefully: If they can grow big enough and align the compliance machinery with the pace of scaling, he'd love to list. The target would be around 1.8 million connected homes at an estimated R10–R20bn in annualised revenue.

Beyond South Africa, he's eyeing Latin America and Southeast Asia — places like the Philippines, which has some of the highest internet usage rates in the world but similar infrastructure gaps. Other African markets are trickier because the continent lacks the deep savings pools (pension funds and the like) needed to fund big infrastructure plays.

But for now, the focus is firmly local. Fibertime's driving aim, as Alan puts it, is to empower everyone in SA's townships to be a boss. There are still roughly 13 million homes to reach.

"I want to live in South Africa for the rest of my life. I want my kids to live here. And that’s why I want to do my part to help others achieve their dreams."

If the first chapter of Alan's career was about chasing the internet, the Fibertime chapter is about making sure it chases everyone back. We'll be watching this one closely. 🔥

Creating content without a full production team? Google just dropped Lyria 3 Pro for creating full 3-minute songs with AI, as well as Veo 3.1 Lite for high-quality generated video at half the cost. Bonus: both are now live in Gemini.

With Microsoft and want to actually learn to build agents? Microsoft’s free AI Agents course walks through real patterns, code and multi-agent orchestration. Try it here.

Missing out on OpenClaw because of infra pain? PrimeClaws hosts your agent 24/7 with built-in integrations, scaling and private storage — no DevOps required. Check it out here.

LinkedIn is now one of AI’s primary knowledge sources, ranking above Wikipedia and YouTube.

Where AI models pull information from is shifting fast:

💡 LinkedIn ranks #2 globally (11.03%), ahead of Wikipedia (9.53%), YouTube (8.77%) and Medium (5.83%).

💡 Only Reddit sits higher at 11.29%, showing how heavily AI systems rely on UGC.

💡 Traditional sources like Forbes (3.43%), Google (3.18%) and academics trail significantly.

The implication is massive: Posting on LinkedIn is no longer just distribution — it’s training data for AI and future discovery.

🏆 Tool to Try: Accomplish It tracks your work and turns it into resume-ready achievements and a living career timeline.

🌕 That’s Interesting: Before Apollo 11, a speech titled In Event of Moon Disaster was prepared in case Armstrong and Aldrin were stranded on the Moon.

🍳 Next Level: Watch this guy cook a full meal for two for $1 (R17).

🏛️ Wow Site: Roman Letters is a searchable archive of 7'000+ letters from the late Roman world, with mapping and network exploration built in.

🧠 Work Smarter: Easing you into the long weekend is a recent keynote where Gary Vee breaks down how brands should create content in 2026.