Metre Lemoisson

@MrGueye_ Et Rust comme language 😂

Un site vitrine et ça utilise kubernetes

@Dominus_Kelvin @oosanya2 @asemota @hagfish_app That's the truth !

@LemoissonMetre @oosanya2 @asemota Yeah getting pricing right is something you have to be intentional about For @hagfish_app I’ve experimented with PPP, and most recently credit based pricing Like why ask a freelancer who doesn’t get gig that often to subscribe to your stuff monthly They will churn

Is there a list of failed venture backed African startups anywhere? I remember someone tried to do one last year, but it wasn't extensive. There are a number of failed YC backed startups not also in the startups.rip list yet. Learning from failure is very important.

@oosanya2 @asemota @Dominus_Kelvin It's not easy to sell an app in our continent 😅

@LemoissonMetre @asemota @Dominus_Kelvin this is embarrassing 😭😂

@asemota @Dominus_Kelvin I tried the subscription model in DRC with my app Dug Assistant. Users started comparing it to rent and most of them could not afford yet another recurring cost. They were also told that we get paid by Telco companies when they buy internet bundles to use our APP

Now I have enough across the site and corroborating sources. Here is the full pattern analysis: The African startup graveyard tells a story that is recognizably similar to the YC graveyard in some places — and starkly different in others. The surface causes (financial distress, market competition, operational failures) are familiar. But the root causes are largely Africa-specific, and they reveal something important about the gap between how the ecosystem has been funded and how African markets actually work. 1. The Silicon Valley playbook applied to markets it was never designed for This is the single biggest meta-pattern. After record funding highs in 2021 and 2022, venture capital inflows into African startups declined sharply by 2024, leading to lower valuations, tighter access to capital, cost-cutting, and a renewed focus on profitability — not because of a lack of innovation, but simply a reckoning with mathematics. Investors funded companies on growth-at-all-costs assumptions borrowed from US tech, in markets with radically different cost structures, consumer purchasing power, and infrastructure realities. iROKOtv spent $100 million trying to win the African streaming market against Netflix and local piracy simultaneously — the platform struggled to compete with global players and ultimately decided to shut down after its subscription model proved unsustainable in markets where most users couldn't or wouldn't pay monthly fees. 2. Currency exposure as an existential threat — not a risk factor This is almost entirely absent from the YC graveyard analysis, but dominates the African one. Okra launched Nebula, a naira-priced cloud offering in October 2024, feeling bold, but it arrived in a market dominated by AWS and Google Cloud — and with the naira losing ground daily, it was a mismatch the company could not win. The same story plays out across fintech, logistics, and e-commerce on the continent: companies earning in local currency while paying for cloud infrastructure, imported hardware, and USD-denominated debt faced a structural arithmetic problem that no amount of product excellence could solve. Naira depreciation alone wiped out the unit economics of dozens of otherwise viable businesses. 3. Financial distress is the headline but governance collapse is the real killer Across the startups that failed, several common themes emerged: financial difficulties at 58%, market-related factors at 20%, regulatory issues at 17%, and operational challenges at 27%. But the data undersells a more disturbing sub-pattern. Bento Africa faced mounting allegations of a tax and pension remittance scam, which led to investigations by the Lagos State Inland Revenue Service and the Economic and Financial Crimes Commission — issues compounded by internal turmoil including the resignation of the CEO and the layoff of the entire engineering team after disputes over unpaid salaries. Thepeer went further: a co-founder went public with allegations of fraud and missing funds, with alleged misuse including $50,000 spent on car purchases for a company generating less than $1,000 in annual revenue. These are not isolated governance failures — they reflect what happens when founder incentive structures, weak board oversight, and a thin investor monitoring culture collide under financial pressure. 4. Aggressive geographic expansion without unit economics — the "Big Four" trap The "Big Four" hubs — Nigeria, Kenya, South Africa, and Egypt — attract almost all the capital, and founders under investor pressure expand across them before proving the model in one. Bento Africa had quickly expanded into Ghana, Kenya, and Rwanda while onboarding hundreds of businesses — it was the kind of company investors expected to scale quickly — but it collapsed in a spiral of mistrust and forensic scrutiny. Lipa Later similarly scaled fast after a large 2022 raise, then accumulated unsustainable debt and made a costly 2023 acquisition, leaving the business overleveraged and entering administration in March 2025. Expansion before consolidation is lethal in markets where each country has different regulatory frameworks, payment rails, tax obligations, and consumer behaviors. 5. Regulatory dependency as a single point of failure KOKO Networks is the most dramatic example on the graveyard: following the government's discontinuation of its carbon credit sales, the Kenyan clean cooking startup laid off all 700 employees and ceased operations, facing bankruptcy after failing to secure government approval to sell its carbon credits. A business model built on a government-administered revenue stream — carbon credits in this case — has no defense when that stream is shut off by policy decision. The same structural fragility appears across the list in fintech (regulatory licensing dependencies), healthtech (government procurement exposure), and agtech (subsidy dependency). African regulatory environments are not just slow and complex; they can reverse course entirely without warning. 6. The open banking / infrastructure pioneer problem Okra built impactful technology, worked with some of the biggest brands across the continent, and helped pioneer open banking in Africa — and still failed. This reflects a pattern specific to infrastructure plays: the pioneer does the expensive work of educating the market, building the regulatory case, and proving the concept, then either gets acquired on poor terms, or finds that incumbents (banks, telcos) simply replicate the capability internally once it's been validated. Okra, Collect Africa, and similar API-layer companies built genuine value but couldn't capture enough of it before the window closed. 7. Pandemic-induced structural damage that never healed Afristay's shutdown closed the chapter on one of South Africa's most recognizable travel-tech platforms. The pandemic crushed demand almost overnight — traffic dropped 96%, bookings evaporated — and by 2023, Afristay was operating with two part-time staff and fewer than 30 monthly bookings. But what makes this distinctly African is the coda: the analysis notes that South Africans now have far less spending power than they did a decade ago, and travel is a luxury many can no longer afford. In many African markets, the pandemic didn't just pause spending — it permanently compressed the consumer base for discretionary digital services. 8. The transparency gap — silent exits deprive the ecosystem of learning Without clear founder communication, the ecosystem couldn't learn from the failure. Silent exits deprive the ecosystem of learning opportunities and damage founder credibility for future ventures. This is the meta-problem that startupgraveyard.africa itself was built to solve. Unlike the YC world where post-mortems are almost a cultural ritual, African startup failures often just disappear — founders move on, companies go dark, employees find out when salaries stop arriving. Medsaf ceased operations in March 2024 but its shutdown only became public knowledge in 2025. Kippa's wind-down was so opaque that by mid-2025 no one could confirm what had actually happened. The fundamental divergence from the YC pattern In the YC graveyard, most companies died from internal causes — wrong timing, feature creep, bad pivots, founder conflict. The ideas were often right; the execution was the problem. In the African graveyard, the pattern is more often external forces colliding with fragile internal foundations: currency collapse, regulatory reversal, funding withdrawal, and infrastructure failure exposing governance gaps that good market conditions had previously masked. The common thread was not simply lack of funding, but fragile unit economics, aggressive expansion strategies, and delayed paths to profitability — 2025 functioned less as a funding winter and more as a reckoning. The African startup that survives is the one that treats macro volatility as a core design constraint from day one — not an afterthought.

Avec mes 8 Go de VRAM, j’ai essayé de petits modèles en offline. Je leur ai tous posé la même question : « Lister les provinces de la RDC ». 😂😂 Il n’y a que Gemma-3-4b qui a essayé d’être correct… Il y a Wayambi, Afristique, etc. Mon préféré : Gemma-3n-e2b-it-text.

@TJMufuta Ils sont vraiment partout

Chaque lundi, Claude fait des carnages 😭

interesting things here : - the ai agent wrote unit tests validating the vulnerable code. 😭😌 - the dev is committing 400+ commits daily, who's reviewing them this fast? what could go wrong? AI usage is fun and all, but remember there's no free lunch, something gotta pay!

1100 copies déjà corrigées avec dugassistant.com. Si vous connaissez un enseignant qui galère encore pour corriger les copies, montrez lui Dug Assistant.

@AleshOfficiel @Tracy_Ntumba Nous offrons aussi le même service avec notre logiciel dugassistant.com. Nous avons lancé cette fonctionnalité il y’a deux semaines et nous avons déjà plus de 1300 copies corrigées avec notre système . Avec des feed-back pour les élèves et des rapports !

Et dire qu’il y a quelques mois on a failli « lapider » la Ministre de l’éducation Nationale en RDC pour ça 😅. Cc @Tracy_Ntumba

Dug Assistant vient de transformer la correction de copies en un véritable jeu d’enfant pour les enseignants. Nous sommes en pleine phase pilote de notre nouvelle version, qui intègre une fonctionnalité de correction automatisée de copies. Désormais, il suffit à l’enseignant de prendre en photo ses questionnaires ainsi que les copies de ses élèves : Dug Assistant s’occupe du reste. Essayez-le gratuitement sur dugassistant.com

I don’t think Hetzner is racist, but their policies are unfairly discriminatory towards Africans. Hetzner was our 1st choice when we wanted to deploy @fansted_ I did KYC, submitted IDs, and was automatically rejected 2 minutes later. Be advised.

“ faut juste demander à ChatGPT , il peut faire ce truc en 2 minutes” , c’est l’attitude des plusieurs clients à l’ère de l’IA. Oubliant que créer des logiciels c est pas que “Coder”. Y’a un tas de processus à faire au préalable

@TheMwami La légèreté…

Le numérique, bien plus que des compétences techniques : stratégies, erreurs à éviter et opportunités au cœur du workshop Talent 4 Startups ce lundi avec @JohnSantos et @LemoissonMetre , un rendez-vous clé pour bâtir votre avenir digital #Talent4Startups #DigitalCareers

Chères autorités est-ce que vos cortèges sont obligés de conduire comme des fous alors que vous avez déjà la priorité en route ? Eh oui! Vos moments de courses effrénés pour nourrir votre adrénaline créent des morts et des vies détruites 💔

Ruth Djamba 26ans a été percutée par un cortège militaire. Comment ceux qui sont censés nous protéger peuvent-ils devenir une menace pour la population ? Pourquoi les ministres, officiers se déplacent en cortège et roulent à une telle vitesse ? Vivre à Kinshasa devient insupportable telle une jungle, l’insécurité est constante et personne ne semble en répondre.

🚨🚨🚨 Ruth Djamba,une autre passagère et leur motard ont été tués par le cortège d’un officier militaire. Demandons justice pour eux💔 #JusticepourRuthDjamba

@TheMwami Surtout quand on fourni pas assez de contexte ou qu'on le laisse générer trop de code à la fois. 😒

Un projet née à l’ISTA Aujourd’hui un produit vendu et installé dans plus de 6 communes. Trust the process 🙌🏾🇨🇩✨

@Olivierkanyind 😂

#Tu es mon employé. Je t'appelle☎️ ta femme, ta fiancée ou ta copine répond au téléphone, je te licencie. #Tu es ma travailleuse je t'appelle ton mari, ton fiancé où ton copin me répond🤙, je te vire également. Je viens de licencier 1.