Feeling down {{ FIRSTNAME | there }}? Well, at least you didn't drive your bakkie off a cliff en route to Cape Point. LOL. Don't worry, the driver's surprisingly unharmed.
Send this to your most Capetonian friends. 🌄
In Today’s Open Letter
Startup Survival: Before you give up, read this story.
Local: May fuel relief & Joburg’s R100M+ founders.
Global: Why Google gave the Pentagon its AI keys.
AI: Anthropic’s $1 trillion & behind SA’s whacky policy.
Chart: Which SA city has the highest-earning citizens?
Work Smarter: Will people actually value your product?
"The Collab helped us land an investor”
Priaash Ramadeen from The Awareness Company is just one SA founder whose business benefited hugely from The Founder Collab.
All 144 members tell similar stories. Don’t build your business alone.
TRENDING NOW
3 Lessons From a Startup That Almost Died
Sometime in 2024, Tim Treagus sat down to draft an article titled "Why We're Shutting Down Yazi…" Two years later, they’re in 15 countries, with clients like Old Mutual, Discovery, Pick n Pay and Capitec, and investors gave them R30 million to grow more – here’s what happened in-between…
The runway was thinning. A fourth funding deal had just slipped. And Meta had killed his Yazi WhatsApp number. Not once or twice, but three separate times.

Every founder, at least 3 times, every single day…
You already know how the story ends: Yazi raised a R30m pre-money round led by 3 Capital Ventures, went live in 15 countries, with corporate clients and a 1.8 million-person panel, 80% of leads now coming through AI search. We covered that in March.
What we didn't cover was the four years before and the lessons Tim took out of them.
1. If you don't build in your area of expertise, expect to pay tuition
Tim started Yazi in 2022 with a finance degree, two years of work behind him building Billo (a small e-commerce brand he'd run since university), and almost nothing else: No domain expertise in research, no network in the industry, no technical background.
"I was paying to learn," he says. Every year in business was an investment in getting to know the buyer, the sales cycle and the language of fundraising. And his co-founder and CTO, Mzwandile Sotsaka, only came on board 18 months in.
Tim’s lesson: Your first years in a new space are about learning. Don’t fight it, budget for it, so you can pursue understanding of that space relentlessly. Oh, and whatever you do, get a technical co-founder in earlier than Tim did.
2. You don’t need to pitch more; you need to be sure what you’re pitching is really ready
Tim started raising in May 2022. What followed was nearly four years of pitch decks, near-misses and deals that fell through at the last minute (on four separate occasions).
The temptation when raising is grinding and assuming the answer is more. More meetings, more intros, more decks out. Now Tim knows better.
The early pitches weren't failing because the deck count was too low. They were failing because the ingredients weren't there yet: the metrics, the panel scale, the story. If you're raising and not getting funding, it’s not about "who else can I pitch?" It's "am I actually ready?"
3. Send the update
This is the unglamorous one. And the one that actually closed the round.
Since 2022, Tim has sent a fortnightly investor newsletter to over 110 people. The numbers, the thinking, the pivots, the bad months. Every two weeks. No exceptions.
It kept him accountable. But more importantly, when 3CV eventually came looking, those 110 people had been watching Yazi happen, one fortnightly email at a time. Every time Tim met a potential investor or industry contact, they went straight onto the list. So by the time a real opportunity came, the relationship was already warm.
The round wasn't won on the pitch; it was won in the previous 90-odd updates.
Oh, and in case you were wondering, the shutdown article never went out.
We're watching this space…
TRENDING IN AI
3 Things in AI this week
The world's most valuable AI company? Anthropic crossed a $1 trillion valuation, celebrating by launching native Claude connectors for Blender, Adobe Creative Cloud, Ableton and Canva. Read more here.
What if your AI assistant could actually see your work? Google's new Workspace Intelligence connects Gemini across Gmail, Docs, Sheets, Slides and Drive, so it builds from your real data, not thin air. See how it works here.
Can we trust any new policy in the age of AI? The government withdrew its own AI policy draft after discovering fictitious references in the document. Praelexis breaks down what went wrong and what needs to change. Read the full analysis here.
Brought to you by Praelexis
With over a decade of experience in AI, Praelexis custom-builds solutions to solve problems in finance, health, agriculture and more.
CHECK THIS OUT

Got a product idea that crosses into hardware?
Most dev shops write code. Octoco's team also solder boards and design PCBs, which is why founders across HealthTech, AgriTech, SecurityTech and FinTech trust them when software alone isn't enough.
Whether you need a fractional CTO to guide your roadmap, a team to engineer a connected device from scratch, or developers who think like engineers to build your platform — start a conversation, not a procurement process.
Based in Stellenbosch and Cape Town, with clients from early-stage startups to established companies. And if your idea is early enough, their Yenza venture studio might even invest in and incubate it.
IN SHORT
Cause dropping a headline or two at the braai never hurt anyone…
⛽ May Petrol Pump Relief. SA motorists can breathe a sigh of relief knowing that the R3 per litre fuel tax relief will officially continue into May, with June and July most likely to still have some relief, albeit at a lower level. Ok, sjoe, ne.
🦄 Joburg’s Got Game. In the last 8 years, 9 Joburg founders built amazing companies and raised a cool R6 billion between them — three of them clocked over a R1bn each. Check out the list of Joburg founders who raised R100M+.
⛏️ Just In The Nickel Time. Patrice Motsepe's African Rainbow Minerals has just signed a deal to sell nickel concentrate from the Nkomati mine to Europe's only large-scale nickel smelter, Harjavalta, in Finland, owned by Swedish Boliden AB.
🤖 Keys To The House. After Anthropic refused to give the US unrestricted AI access without guardrails against mass surveillance and autonomous weapons, Google gave the DoD access to all of its AI for classified networks. There's always someone willing…
🎯 Leads Unlocked. Former Springbok Bob Skinstad has the presence but not the time to build LinkedIn. So FDC turned his profile into a lead engine that's driven 33.5 million impressions, 70k+ new followers and at least 2 quality business leads per month.*
* From our partners. Find all the best service providers for your business in our Founder Stack.
WORK SMARTER

How to be sure people will buy your product BEFORE you spend time/money making it…
How to write a value hypothesis before you build your product
Most founders believe in their idea so much that they’ll spend months or years building something no one is going to buy. To save you from the cost and heartache of building the wrong thing, we got an SA UX specialist in to tell us how they ensure they always build the right product or feature next.
Turns out there are 5 specific steps to follow, and it all centres around knowing how to write a value hypothesis before you start building.
WHAT YOU SAID
And we have lift-off…
Yesterday, we showed you how Launch Oracle builds early traction, asking about your biggest launch challenge. Most say it’s in validation…
🟨🟨⬜️⬜️⬜️⬜️ 🎯 Building an audience before launch day (17%)
🟩🟩🟩🟩🟩🟩 💸 Validating demand without spending a fortune (44%)
⬜️⬜️⬜️⬜️⬜️⬜️ 📣 Cutting through the noise on launch day (0%)
🟨🟨🟨⬜️⬜️⬜️ 🔁 Converting interest into actual sales (24%)
🟨🟨⬜️⬜️⬜️⬜️ 🤷 Honestly, I'm figuring it out as I go (15%)
Your 2 cents…
“I know we have an absolutely amazing product to solve for unemployment but uptake is very slow.”
Interesting, Thando! Question: Who does the product really benefit, versus who pays for it? Those need to match 100%; otherwise, you battle to sell. You’re welcome to share more about your product privately with the team, if you like… 🌈
UP OR DOWN?
A graph that matters
Joburg earns the most per taxpayer — but Cape Town is growing its tax base faster.
💡 Johannesburg's average taxable income is R109'000 more than Cape Town's and R68'200 more than Tshwane's — the widest metro gap in the country.
💡 Despite a slightly larger population, Joburg has fewer assessed taxpayers than Cape Town: 970'892 vs 975'779 in 2024.
💡 Cape Town's taxpayer base grew 57% over the last decade vs Joburg's 40%, though much of that growth reflects SARS bulk-registration drives, not new economic activity alone.
The income gap says Joburg still drives SA's highest-value economy, but the growth gap says the talent and tax base are slowly shifting south.
AROUND THE WEB
The most fun today…
🧩 Tool to Try: TransVoix lets you call anyone, speak in your language, and they hear you in theirs, in your actual voice, in under a second.
🌍 That's Interesting: 2.4 billion years ago, the evolution of oxygen-producing bacteria caused a mass extinction: Oxygen was toxic to existing life, and its reaction with methane triggered a 300-million-year-long ice age.
🎳 Next Level: Watch the world's fastest strike in ten-pin bowling at 225 km/h.
🌍 Wow Site: LexiExplorer lets you learn languages by exploring free 3D environments.
NEXT STEPS
Here’s how to get more
Join our online community built for founders and startup/tech enthusiasts called The Founder Collab.
Vote in the poll below and leave a comment.
How are you feeling about today’s Open Letter?




