Audit Zendesk Ticket SLA Breaches and Resolution Time from CSV

How to export Zendesk ticket metrics to CSV

Zendesk provides two ways to export ticket-level duration metrics to CSV: a full account ticket CSV export from Admin Center, or an Explore dataset export from the Support - Tickets dataset.

1. From Zendesk Admin Center (Full CSV export): Go to Admin Center > Account > Tools > Reports, locate the Full CSV export option (available on Support Professional, Enterprise, and Suite plans when enabled, as of October 2026), and download the ticket CSV.
2. From Zendesk Explore: Open a table report in the Support: Tickets [default] dataset containing ticket ID, timestamps, and duration metrics, then click Export > CSV.
3. Drop the downloaded .csv file into xmetric (the app also imports .xlsx).

Three support SLA and resolution formulas and SQL equivalents

1. SLA Breach Rate by Priority Tier

Even if a ticket was created before an official SLA policy was attached in Zendesk, you can test every historical ticket against your target response thresholds (for example: Urgent <= 60 business minutes, High <= 240 minutes, Normal <= 480 minutes).

xmetric / Tableau formula ([First Reply SLA Status]):

IF [Priority] = 'Urgent' AND [First reply time in minutes within business hours] > 60 THEN 'Breached'
ELSEIF [Priority] = 'High' AND [First reply time in minutes within business hours] > 240 THEN 'Breached'
ELSEIF [Priority] = 'Normal' AND [First reply time in minutes within business hours] > 480 THEN 'Breached'
ELSE 'Met SLA'
END

SQL equivalent:

SELECT
  "Priority" AS priority,
  "Group" AS support_group,
  COUNT(*) AS total_tickets,
  ROUND(MEDIAN("First reply time in minutes within business hours"), 1) AS p50_first_reply_mins,
  ROUND(QUANTILE_CONT("First reply time in minutes within business hours", 0.90), 1) AS p90_first_reply_mins,
  ROUND(MEDIAN("Full resolution time in minutes" / 60.0), 1) AS p50_full_resolution_hours,
  ROUND(
    COUNT(*) FILTER (
      WHERE ("Priority" = 'Urgent' AND "First reply time in minutes within business hours" > 60)
         OR ("Priority" = 'High' AND "First reply time in minutes within business hours" > 240)
         OR ("Priority" = 'Normal' AND "First reply time in minutes within business hours" > 480)
    ) * 100.0 / NULLIF(COUNT(*), 0),
    1
  ) AS sla_breach_rate_pct
FROM zendesk_tickets
WHERE "First reply time in minutes within business hours" IS NOT NULL
GROUP BY 1, 2
ORDER BY total_tickets DESC;

2. One-Touch Ticket Rate vs. Reopen Rate

Agents who rush to mark tickets Solved on the first reply to beat resolution targets often suffer high Reopens. Plotting One-Touch Rate (Replies = 1 AND Reopens = 0) next to Reopen Rate (Reopens >= 1) keeps speed incentives honest.

xmetric / Tableau formula ([One-Touch Resolution Rate]):

SUM(IF [Replies] = 1 AND [Reopens] = 0 AND [Status] IN ('Solved', 'Closed') THEN 1 ELSE 0 END)
/
NULLIF(SUM(IF [Status] IN ('Solved', 'Closed') THEN 1 ELSE 0 END), 0)

3. Ticket Creation Heatmap by Hour of Day and Day of Week (FIXED LOD)

Staff your support schedule where ticket volume and SLA breaches actually spike by comparing each (Day of Week, Hour) slot against your weekly average hourly intake.

xmetric / Tableau formula:

COUNTD([Id]) / MIN({ FIXED DATEPART('weekday', [Created at]) : COUNTD([Id]) })

Limits of Zendesk CSV exports

• First Resolution vs. Full Resolution: In Zendesk, First resolution time in minutes stops the clock the first time an agent marks the ticket Solved, whereas Full resolution time in minutes includes time after the customer replies and reopens the ticket. Always check Full resolution time in minutes when customers complain about slow outcomes.
• Auto-closed system tickets: Automated webhook notifications or spam tickets closed without a public agent reply have a blank First reply time in minutes. Filter out tickets where Via = API (or tag contains auto_close) so system noise does not distort your human support metrics.