Monday’s network outage at Amazon Web Services took down a third of the internet, disrupting everything from financial platforms to AI applications. The downtime at the world’s largest cloud provider, traced to a regional gateway issue on the US East Coast, has been resolved, but it left businesses questioning the real cost of relying on a single cloud provider and the price of ensuring that it never happens again.
Two Web3 startups, Azza, a Nigerian social commerce crypto app, and AZA Finance, a cross-border payments platform reportedly in talks to be acquired by dLocal, a Latin American fintech unicorn, were among those affected by the outage. Both experienced temporary disruptions to their transaction processing and user dashboards, underscoring how even blockchain-driven firms remain vulnerable to failures in centralised infrastructure.
The outage highlighted a growing dilemma for startups that depend on AWS and similar services. Building a redundant, multi-cloud server system ensures uptime but eats into already tight operational budgets. The trade-off between resilience and cost is becoming one of the most pressing decisions for early-stage tech startups.
Palremit, a Nigerian crypto offramp platform, stayed online throughout the outage by routing its servers across multiple AWS regions.
“If you build a redundant system, it automatically switches to another server anytime there’s a downtime in one region,” said CEO Ugochukwu Mbamalu. “That’s how you ensure uptime.”
Mbamalu said the redundant setup costs about $10 extra monthly, but this cost can easily grow as the startup achieves scale. Palremit currently spends hundreds of dollars each month on cloud services for user data management and other database operations.
“I was affected by a cloud outage once,” said Mbamalu. “After that, I decided to hedge my bets on multiple server options.”
Azawire, a Nigerian stablecoin payments startup, avoided disruption because its AWS data centres are based in Frankfurt, Germany.
“[The AWS downtime] was a regional issue,” said Emmanuel Onyo, CEO of Azawire. “Servers that were out should be the US-based data centres. Azawire is hosted in Frankfurt, so we didn’t experience much of a lag.”
James Lucky, a Web3 software developer and CTO of GrowSafe, a Nigerian crypto swapping platform, said the startup’s infrastructure was largely unaffected because he hosts most of his projects on DigitalOcean.
“Most of my servers are on DigitalOcean and other providers, so I wasn’t really affected,” said Lucky. “But my apps on Vercel were down, and it made me suspect Vercel is powered by AWS.”
James added that DigitalOcean, a US-based cloud services provider which holds 2% of the market, remains a favourite for his operations as an early-stage startup because of its low-cost—compared to AWS—flexibility, and deployment speed.
“You can host apps on Digital Ocean for at least $5,” said Lucky. “And you can host all kinds of apps, from frontend to backend services, and cron jobs. I’ve never had downtime on DigitalOcean.”
Adaobi Orajiaku, CEO of Atsur, a Nigerian Web3 art-tech startup, said her AWS-hosted system also stayed up because it wasn’t powering any heavy operation during the time of the outage.
“We [Atsur] didn’t have any onsite archiving or verification work [going on] at the time,” said Orajiaku. “Else we would definitely be having issues and may just have to resort to cataloguing on spreadsheets, and postpone the actual verification [of artwork] and/or entry.”
Customers who experience issues with Atsur’s app typically reach out directly to the team, but so far there have been no reports of disruption. The startup said it is keeping an eye on performance and checking in with B2B gallery users to ensure everything remains stable, said Orajiaku.
It is somewhat of an irony that the decentralised technology that Web3 startups champion still depends on centralised infrastructure. Blockchain ensures continuous transaction flows with maximum uptime, yet user data and platform management rely on centralised cloud providers like Amazon, Google, and Microsoft. The contradiction creates friction each time the backbone of the internet falters, yet for most startups, there is no real alternative.
“These are not common outages that happen all the time,” said Onyo. “If global, bigger companies like Perplexity and Coinbase did not think of having duplicate servers with other cloud platforms, then startups will obviously struggle too.”
