The longest single incident StatusBird recorded in the last 90 days was Plaid, which ran 1322 minutes from August 2, 2026 at 14:52 UTC until recovery on August 3, 2026 at 12:54 UTC. The next four longest were Shippo at 560 minutes, ShipStation at 254 minutes, Shippo again at 248 minutes, and Shippo a third time at 216 minutes, which means one payments service and two shipping services account for every one of the top five.
StatusBird checks each of the 84 services it tracks every 2 minutes, and every figure below comes from those readings over the 90-day window ending September 25, 2026. The methodology page explains how a run of non-operational readings becomes an incident.
Which incidents were the longest?
| Service | Start (UTC) | Recovered (UTC) | Duration | Worst status |
|---|---|---|---|---|
| Plaid | August 2, 2026 at 14:52 | August 3, 2026 at 12:54 | 1322 minutes | major |
| Shippo | July 30, 2026 at 04:42 | July 30, 2026 at 14:02 | 560 minutes | major |
| ShipStation | September 25, 2026 at 07:57 | September 25, 2026 at 12:11 | 254 minutes | major |
| Shippo | August 7, 2026 at 14:18 | August 7, 2026 at 18:26 | 248 minutes | major |
| Shippo | July 14, 2026 at 07:21 | July 14, 2026 at 10:57 | 216 minutes | major |
Plaid, August 2 to 3: 1322 minutes across two calendar days
This one is the outlier of the window. StatusBird's first non-operational reading came at 14:52 UTC on August 2, 2026, and the service did not return a clean reading until 12:54 UTC the following day: 1322 minutes, worst status major. No other incident in the 90-day window came close to half that length.
The cross-midnight shape is what made it dangerous. An incident that opens mid-afternoon UTC and closes late morning the next day covers a full overnight period in every timezone, so a store owner who checks dashboards at the start of the working day would have found the failure already 20 or so hours old. If your checkout offers bank-account payment or you use account linking for customer verification, every attempt during that window was hitting a service StatusBird could not get an operational reading from. The practical consequence is not a visible error page on your storefront, it is a payment method that silently fails while card options keep working, which customers rarely report and analytics rarely flag.
Plaid finished the 90 days on 98.8% uptime with grade C, 4 incidents rated major or critical, and 1548 minutes of total downtime. This single incident is 1322 minutes of that 1548 minutes. Its second-longest, on July 10, 2026, ran 198 minutes from 16:06 UTC to 19:24 UTC. Plaid's average incident length of 387 minutes is the highest of any service in the table, and that average is essentially one very bad day pulling everything up. The full record is on the Plaid reliability history.
Shippo, July 30: 560 minutes inside one business day
StatusBird logged Shippo as non-operational from 04:42 UTC to 14:02 UTC on July 30, 2026, a run of 560 minutes with worst status major. That window opens before the European working day and closes in the middle of the US morning, so it overlaps the label-printing window for stores on both sides of the Atlantic.
A store that generates labels through Shippo would have seen the day's pick-and-pack queue stall while orders kept arriving. Rate lookups are the second-order problem: if your checkout asks a shipping provider for live rates, a provider that cannot answer means the rate step either falls back to your flat rates or shows nothing at all. Nine hours of that is enough to miss a carrier pickup cutoff, which pushes an entire day of orders into the next dispatch.
ShipStation, September 25: 254 minutes this morning
The third-longest incident is also the most recent: ShipStation was non-operational from 07:57 UTC to 12:11 UTC on September 25, 2026, lasting 254 minutes at major severity. That is over four hours of a single morning.
ShipStation's 90-day numbers are better than its incident lengths suggest: 99.67% uptime, grade B+, 420 minutes of downtime, and only 2 incidents rated major or critical. The other one ran 166 minutes on June 30, 2026 from 16:43 UTC to 19:29 UTC. So ShipStation fails rarely, but when it does it averages 210 minutes. For a store, that profile means you should not plan around ShipStation being briefly unavailable. Plan around it being unavailable for the length of a shift. Current readings are on the ShipStation status page.
Shippo, August 7 and July 14: 248 minutes and 216 minutes
Shippo takes the fourth and fifth slots too. On August 7, 2026 it was down from 14:18 UTC to 18:26 UTC for 248 minutes, and on July 14, 2026 from 07:21 UTC to 10:57 UTC for 216 minutes. Both were major. A fourth Shippo incident, 198 minutes on August 13, 2026 from 12:52 UTC to 16:10 UTC, only just missed the top five.
Across the window Shippo recorded 98.93% uptime, grade C, 12 incidents rated major or critical, and 1382 minutes of downtime, with an average incident of 115 min. That is a different failure shape from ShipStation: more frequent, individually shorter, and spread across the calendar. Its most recent incidents were much briefer, 30 minutes on September 24, 2026 starting 19:25 UTC and 8 minutes the same day starting 08:52 UTC, and StatusBird currently models a 15.0% chance of a non-operational Shippo reading in the next 24 hours. The Shippo reliability history lists each one.
What do these five have in common?
Three things stand out from the timeline data:
- Shipping and payments carried the window. Four of the five longest incidents belong to Shippo or ShipStation, and the longest belongs to Plaid. Stripe, PayPal, Square, Adyen, Braintree, Klarna, and Authorize.net all finished the 90 days at 100.0% uptime with 0 incidents.
- One incident can define a month. August 2026 recorded 16 major or critical incidents and 40.4 hours of combined downtime across 9 services, and Plaid was the most affected service that month.
- Long incidents start early or cross midnight. Three of the five opened between 04:42 UTC and 14:52 UTC, and the Plaid one ran through the night.
What should a store do with these numbers?
- Decide your fallback per dependency, in advance. For a 560 minute label outage that means a written answer to "which carrier portal do we log into directly, and who has the credentials." For a 1322 minute bank-linking outage it means knowing which alternate payment method you switch on and who flips it.
- Set your shipping rate fallback now. If your checkout calls a live rate API, configure flat-rate fallback rates so a provider outage does not leave the rate step blank.
- Route overnight alerts to SMS, not email. The Plaid incident opened at 14:52 UTC and was still unresolved at breakfast the next morning. An email that lands in an inbox nobody reads at 03:00 does not shorten the window.
- Reconcile against the exact recovery time. ShipStation recovered at 12:11 UTC on September 25, 2026. Orders created before that time are the ones to re-check for sync, not a vague "this morning."
- Check the frequency profile before you sign. Shippo's 12 incidents and ShipStation's 2 incidents produced similar-looking bad days for very different reasons. The outage index shows the month-by-month totals.
If you want to know when one of these services stops responding rather than finding out from a customer email, StatusBird checks each of them every 2 minutes and sends alerts by SMS, email, Slack, Teams, or Discord. You can create an account and pick which services matter to your store.
FAQ
What was the longest outage StatusBird recorded in the last 90 days?
Plaid, at 1322 minutes. It started August 2, 2026 at 14:52 UTC and recovered August 3, 2026 at 12:54 UTC, with a worst observed status of major.
Which service had the most long incidents?
Shippo, which holds three of the five longest incidents in the window at 560 minutes, 248 minutes, and 216 minutes. Over the full 90 days it recorded 12 incidents rated major or critical and 1382 minutes of downtime.
Is ShipStation less reliable than Shippo?
On these readings, no. ShipStation recorded 99.67% uptime and 420 minutes of downtime from 2 incidents, while Shippo recorded 98.93% and 1382 minutes from 12 incidents. ShipStation's incidents are longer on average, at 210 minutes each.
How severe were these five incidents?
All five had a worst observed status of major. None of the top five reached critical in StatusBird's readings.
Did any month look worse than August 2026?
Yes. April 2026 recorded 20 major or critical incidents and 122.1 hours of combined downtime across 11 services, compared with 16 incidents and 40.4 hours in August 2026.