Plaid had a major outage on September 26, 2026 that lasted about 1 hour 34 minutes. StatusBird's own monitoring recorded the first non-operational reading at 03:41 UTC and the first clear reading at 05:16 UTC, a measured window of 94 minutes.
StatusBird checks Plaid every 2 minutes from its own infrastructure, so the timeline below is what an independent observer saw rather than what the vendor's status page said afterwards. Every figure on this page comes from those readings.
When did Plaid go down and when did it recover?
| Event | Observation |
|---|---|
| First non-operational reading | 03:41 UTC, September 26, 2026 |
| Worst status observed | major |
| Recovery observed | 05:16 UTC, September 26, 2026 |
| Duration | 94 minutes |
| Non-operational readings | 47 |
Across the window StatusBird recorded 47 non-operational readings (47 major). Readings are taken every 2 minutes, so the true start and end may sit up to 2 minutes either side of the times above. All times are UTC.
How unusual is this for Plaid?
In the last 90 days StatusBird recorded 5 major or critical Plaid incidents, 1640 minutes of combined downtime, and 98.73% availability. That makes this incident shorter than the 328 minute average for Plaid over the same period.
| Date | Started | Duration | Worst status |
|---|---|---|---|
| August 2, 2026 | 14:52 UTC | 1322 minutes | major |
| July 16, 2026 | 09:22 UTC | 8 minutes | major |
| July 11, 2026 | 01:56 UTC | 20 minutes | major |
| July 10, 2026 | 16:06 UTC | 198 minutes | major |
Each of those windows was measured the same way as the one above, from the first non-operational reading to the first reading that cleared it. The full record, including the reliability grade, is on the Plaid reliability page.
What does a Plaid outage affect?
For online stores using Plaid, checkout payments can fail or hang, which means lost orders for every minute of downtime. The size of the loss depends on when the window landed: 94 minutes across a weekday peak is a different number from the same window overnight, which is why the start time above matters more than the duration alone.
What should a store do when Plaid goes down?
The response that matters happens in the first few minutes, before the vendor has posted anything on its own status page. In order:
- Check whether orders are still being created without a captured payment. Those are the ones that quietly go missing once service returns.
- If you run a second processor, switch the default gateway at checkout rather than leaving customers on a failing one.
- Turn off paid traffic to product and checkout pages. Spend keeps running while conversion is zero.
- Pull the list of orders attempted during the outage window and reconcile authorizations against captures once the processor recovers.
- Confirm the problem is the vendor and not your own integration, by checking whether the failure affects every request or only yours.
- Tell your support team what is happening and what to say, before the tickets arrive rather than after.
- Pause any automated job that retries against the failing service, so a backlog does not fire twice on recovery.
- Write down the start time. Most vendor SLA credit claims require the window, and reconstructing it later from memory rarely works.
The longer version, with the checks specific to this service, is in what to do when Plaid goes down.
Knowing before your customers do
StatusBird detected this incident automatically and alerted affected store owners by SMS, email, and Slack within minutes of the first non-operational reading. It monitors Plaid alongside the other third-party services online stores depend on, and paid plans watch your own storefront as well. Start monitoring free.
Frequently asked questions
How long was the Plaid outage on September 26, 2026?
About 1 hour 34 minutes. StatusBird observed the first non-operational reading at 03:41 UTC and the first clear reading at 05:16 UTC, for a measured window of 94 minutes.
How bad did it get?
The worst status observed during the window was major. StatusBird normalizes every vendor status page onto the same four levels, operational, degraded, major and critical, so that severity means the same thing across services.
How often does Plaid go down?
In the last 90 days StatusBird recorded 5 major or critical Plaid incidents and 98.73% availability, with the average incident running 328 minutes. The current figure is always on the Plaid reliability page.
Was my store affected?
If your store calls Plaid during checkout or fulfillment, requests made between 03:41 UTC and 05:16 UTC on September 26, 2026 are the ones to check. Orders created in that window without a matching downstream record are the usual casualty.
Where does this data come from?
StatusBird checks Plaid every 2 minutes from its own infrastructure and records each reading. These figures are those readings, not a restatement of the vendor's own incident report. The method is documented at statusbird.io/methodology, and live status is at statusbird.io/status/plaid.