Methods, apparatuses and systems for diagnosing network anomalies and allocating repair resources in a computer network receive network anomaly reports (NARs) from a group of customers. Each NAR specifies a network anomaly observed by a customer at a network endpoint. Each NAR is entered into a database and associated in the database with attributes including at least one customer attribute associated with the sender of that NAR or network attribute associated with the network anomaly specified in that NAR. A time period is detected during which the difference between an aggregate number of NARs received within that time period and a baseline number of NARs received during prior time periods of substantially identical duration exceeds a threshold. The attributes associated with the NARs received during the time period are analyzed to identify a network anomaly specified in those NARs. Repair resources are further preferentially allocated to the identified network anomaly.