Our team were alerted by a small group of customers that our API had some degradation between 5:17PM and 5:29PM AEDT. This has been isolated to a scaling issue and resolved, most online orders would have succeeded with a retry during this time.