Clanker CloudA support report can be wrong even when its theme names sound right. Reopened tickets may be counted twice, blank categories may disappear, and several messages from one customer may look like several separate incidents.
Use Clanker Cloud Workspace to prepare a trend report from a dated, sanitized export. Define the unit being counted and ask for a supporting table. Check the numbers before turning a pattern into a product priority.
Define the reporting question
Are you counting tickets created, conversations closed, or customers reporting a problem? Those are different measures. Specify the date field, period, and timezone that define the selection.
Keep only the fields needed for the analysis. A ticket ID, created date, category, status, and sanitized issue summary may be enough. Avoid uploading unnecessary customer details or credentials.
Test the counting rules on a small sample
Use a fictional packet before a large export:
| Ticket | Supplied status | Supplied category |
|---|---|---|
| T-101 | Closed | Login |
| T-102 | Reopened | Upload |
| T-103 | Open | Blank |
| T-102 | Reopened | Upload |
Under a rule of one record per ticket ID, this packet contains three unique tickets and one duplicate row. The missing category remains unknown. It should not be assigned to the largest group simply to complete the table.
Request the report and its evidence
Analyze this fictional ticket export for the stated period.
Count unique ticket IDs and list duplicate rows separately.
Keep missing categories in an unknown group.
Save support-trends.md and ticket-count-checks.csv.
Include counting rules, category counts, evidence IDs,
limitations, and questions for the support lead.
Do not infer root causes, close tickets, or send replies.
Ask the Agent to distinguish a supplied category from a category it proposes based on the summary. Review proposed labels before using them in a comparison across periods.
Check the denominator before interpreting a trend
Open the saved table and verify the unique-ticket count. Reconcile the category totals, including unknowns, to that count. Inspect the records behind a prominent theme.
If you compare months, confirm that the same date field, duplicate rule, and category definitions apply. A new tagging policy can change the chart without showing a change in customer experience.
Treat explanations as hypotheses unless the source packet supports them. More upload-related tickets does not by itself establish a deployment defect or identify which customers were affected.
Keep the result useful for the next review
Save the checked counting rules with the Project instructions. Retain the export date and the report’s limitations. A saved CSV is a snapshot; repeating the analysis still requires a fresh export or a verified connection.
Use the feedback analysis starter for a broader qualitative review. Keep proposed actions in a separate section so they do not get mistaken for work already completed.
When a theme needs a case-level handoff, prepare a support escalation brief. Carry the relevant ticket evidence into its timeline and distinguish reported symptoms, suggested steps, and confirmed findings.
Common questions
Does uploading a ticket CSV connect my help desk?
No. It supplies a snapshot. Direct ticket access or changes require a supported connection and appropriate permissions.
Can the report prove why support demand increased?
An export may show a change under defined counting rules. Establishing its cause usually requires additional evidence. Keep the explanation provisional until verified.
Sources
Give your assistant one useful job
Build a source-linked summary with counting rules, examples, and questions for review.