Ticket topics
In short: The Ticket Topic Analytics report shows what customers write about. aishaa reads and understands every customer message and places it on one of the organisation’s topics. The report counts those messages per topic and shows the distribution, the top topics with their subtopics and the trend per topic.
Where to find this
Go to Analytics > Topics. The page is called Ticket Topic Analytics with the subtitle Explore what customers talk about.
Who: from the Team lead role. For accounts with several organisations, you choose with Organisation which ones are included; if you choose nothing, all organisations you have access to are included.
Note: this is not the same as the Topics item in the Organisation group of the menu. There an Organisation admin manages the topics themselves and what aishaa knows about them.
The Ticket Topic Analytics report (menu Analytics > Topics) with the filters, the five cards, the Topic distribution donut and the Top topics table.
Where the topics come from
The topics are the organisation’s topics and subtopics: the same structure you see in the taxonomy explorer and on What aishaa knows. See Taxonomy: support areas and query patterns.
- Every customer message counts once, for the topic of the main question in that message.
- Only messages from customers count, at the moment they came in. The Channel filter looks at the channel of the message.
- A message that aishaa understood but did not place on any topic, or not surely enough (see Minimum certainty), counts as No topic.
- A message that aishaa has not read (for example from a customer in a customer group with AI analysis switched off) only counts in Customer messages.
Filters
Choose the filters and click Refresh to load the report.
| Filter | Meaning | Default |
|---|---|---|
| Date range | The period. Quick buttons 7d, 30d, 90d, 180d and YTD, or your own start and end date. | Last 90 days |
| Group by | The size of the time periods in the trend charts: Auto, None, Daily, Weekly or Monthly. With Auto, aishaa chooses daily up to 31 days, weekly up to 180 days and monthly above that. Daily is disabled above 120 days. | Auto |
| Organisation | Only for accounts with more than one organisation. | All organisations you have access to |
| Channel | Only messages from the chosen channels. | All channels |
| Top N topics | How many topics the donut and the table show separately (0 to 20). The rest falls under Other topics. | 8 |
| Max buckets | An upper limit for the number of time periods (60 to 400). If there would be more, aishaa merges time periods. | 180 |
| Max series | How many topic lines the trend chart shows (1 to 12), never more than Top N topics. | 8 |
| Show “No topic” | Show understood messages without a topic as their own slice and trend line. | On |
| Show subtopics | List the subtopics under each topic in the table. | On |
| Minimum certainty | Only count a topic when aishaa was at least this sure, on a scale from 0 to 1. Less sure messages count as No topic. Slider in steps of 0.05. | 0.00 |
aishaa remembers per user the date range (quick buttons only), the grouping and the chosen organisations. That choice also applies to the Response times and Volume reports.
The five cards
| Card | Meaning |
|---|---|
| Customer messages | Messages from customers in the selected period and channels. |
| Understood | Customer messages aishaa has read and understood. Shares on this page are shares of these messages. |
| On a topic | Understood messages that aishaa placed on one of your topics. |
| No topic | Understood messages that fit none of your topics, or not surely enough. |
| Bucket | The grouping used, for example Daily, or Weekly ×2 if aishaa has merged time periods. |
Charts and table
Below the cards are two parts about the distribution and two trend charts. They use the same filters and period as the cards.
Topic distribution
The Topic distribution donut (Top topics) shows the number of Understood messages in the middle. The slices are the top topics, then Other topics and, if Show “No topic” is on, No topic. Each slice shows the number and the percentage. The percentages are shares of the understood messages, not of all customer messages.
Top topics
The Top topics table (Topic distribution) has the columns Topic, Count and Share (to one decimal place). A topic has the same colour as in the donut. If Show subtopics is on, the subtopics are indented under their topic. At the bottom are Other topics and No topic.
Trend
The Trend line chart (with the grouping in the title) has the subtitle How selected topics evolve over time. There is one line per topic, an Other topics line and, if it is on, a No topic line. The colours are the same as in the donut. Click a topic in the legend to hide or show the line.
Overall volume
The Overall volume line chart (Customer messages) shows the number of customer messages per time period, including messages that aishaa has not read.
With None there are no trend charts; you then see No trend data for selected range.
Good to know
- The shares are shares of understood messages. A customer who writes often therefore weighs more heavily than a customer who sends one message.
- The report uses the current names of the topics. If you give a topic a different name, that name also changes in earlier periods.
- A frequently used No topic can mean that the organisation’s topics do not match what customers write about. In that case, look at the Taxonomy explorer.
- If loading fails, it says Failed to load topic analytics. If there is no data, it says No topic data.
- There is no export of this report. See Sharing figures without an export.
