We’re often told the key to success with AI is having clean data. The AWS outage this week should remind us of the importance of resilient data as well.

The AWS outage interrupted people trying to work, play games, watch other people play games, shop and even eat. That’s because it took down corporate networks, gaming sites, pro sports ticketing, retail stores and restaurants – proving once again how much everyone relies on cloud services.

“When AWS sneezes, half the internet catches the flu,” said Monica Eaton, founder and CEO of Chargebacks911. “Outages like this cause frustrated users, but also triggers a domino effect across payment flows. Failed authorizations, duplicate charges, broken confirmation pages, all of that fuels a wave of disputes that merchants will be cleaning up for weeks. And once a customer files a dispute, you are already on the back foot.”

By now, it’s obvious that these outages will happen again. They’ve happened before to AWS and other clouds, and this recent problem came during a minor DNS update, not a rare event like weather or a cyberattack. The lesson here is: always be prepared for another outage as you would prepare for any other disaster.

"The real story isn't just that AWS had a critical issue, but how many businesses discovered their platform partner had no plan for it, especially outside of U.S. hours," Ismael Wrixen, CEO of software developer ThriveCard, told the National, a site based in the Middle East. "This is a harsh wake-up call about the critical need for multi-regional redundancy and intelligent architecture."

Data resiliency requires a combination of tactics, such as having multiple recovery sites and clouds, the ability to fail over across regions, and perhaps operating in a hybrid environment. Redundancy is a key strategy.

“Based on what we are observing, the most immediate and profound impact of this AWS outage will be a substantial, if reluctant, acceleration toward true multi-region and potentially multi-cloud architectures,” HyperFrame Research analysts Ron Westfall and Steven Dickens wrote in a research note about the AWS outage. “This is not about leaving the public cloud, but about mitigating its intrinsic centralization risk. The scale of the disruption has made the total cost of downtime far too high for companies like Venmo, Snapchat, and the many others impacted.

”A multi-cloud strategy, which involves distributing critical workloads across providers such as AWS, Azure, Google Cloud, and OCI has long been discussed but often avoided due to complexity and cost. Now, it moves from a 'nice to have' to a 'must have' for companies with a genuinely global and always on mandate.”