SaaS Integration Escalation Sample Dataset

By TechSupport Team. Published November 20, 2026. Updated .

Use this synthetic 12-ticket dataset to practice auditing an L2/L3 integration queue. It contains fictional cases for orders, inventory and shipment processing. The sample is intended for training, spreadsheet setup and evaluation of a candidate support team.

Suggested citation · TechSupport Team, “SaaS Integration Escalation Sample Dataset,” version 1.0. Version 1.0. Reuse under CC BY 4.0.

Copy the sample CSV

Twelve synthetic support tickets, version 1.0
case_idflowfault_classwait_minproduct_eng_minl3_fixreopened
DEMO-01ordermapping4060yesno
DEMO-02orderduplicate250nono
DEMO-03orderauth1530noyes
DEMO-04ordertimeout5545yesyes
DEMO-05inventoryschedule8075yesyes
DEMO-06inventoryschedule300nono
DEMO-07inventoryvalidation6030nono
DEMO-08inventorytimeout500nono
DEMO-09shippingmapping2030yesno
DEMO-10shippingmapping3545yesyes
DEMO-11shippingpartner650nono
DEMO-12shippingtimeout4515nono

wait_min is the time to first useful evidence in minutes. product_eng_min is product engineering time in minutes. l3_fix and reopened are yes or no.

Sample escalation CSV

case_id,flow,fault_class,wait_min,product_eng_min,l3_fix,reopened
DEMO-01,order,mapping,40,60,yes,no
DEMO-02,order,duplicate,25,0,no,no
DEMO-03,order,auth,15,30,no,yes
DEMO-04,order,timeout,55,45,yes,yes
DEMO-05,inventory,schedule,80,75,yes,yes
DEMO-06,inventory,schedule,30,0,no,no
DEMO-07,inventory,validation,60,30,no,no
DEMO-08,inventory,timeout,50,0,no,no
DEMO-09,shipping,mapping,20,30,yes,no
DEMO-10,shipping,mapping,35,45,yes,yes
DEMO-11,shipping,partner,65,0,no,no
DEMO-12,shipping,timeout,45,15,no,no

Check your analysis

Median time to first useful evidence. Sort wait_min and average the sixth and seventh values. (40 + 45) ÷ 2 = 42.5 minutes.

Product engineering time. Sum product_eng_min. The sample totals 330 minutes or 5.5 hours. Mapping accounts for 135 minutes across three cases.

Reopened cases. Count yes in reopened. Four of 12 cases were reopened, or 33.3% of the sample. Five cases required an L3 fix.

For a real queue, define a useful evidence timestamp, record exposure to each incident class and review comparable periods. The figures above describe only the fictional sample.

Get help with your support queue

Source and reuse

Credit TechSupport Team and link to this page when sharing or adapting version 1.0 under CC BY 4.0.