A user guide: Discovery tab
Summary
The Discovery tab provides summarised insights into your customer conversations. It includes a detailed report with analytics and AI-driven data on your contacts, helping you identify key patterns and trends. To enhance usability, the feature offers various customisation options, allowing you to tailor how data is displayed according to your specific requirements. This guide walks you through accessing, understanding, and customising the insights available on the Discovery tab.
Using the insight report
Navigation Prompt
Imported Contacts & filters > Discovery tab
Within this tab you'll find key visualisation charts and a table with summarised insight data. You can customise the data to view the most relevant information for your analysis.
Customise your data
Imported contacts filters
The charts and data table interact with the imported contacts filters. You can use the filters to narrow down the sample of conversations that you'd like to see insight information for.
Grouping data
To easily identify trends and patterns you can customise how the information is grouped. Select a grouping option from the dropdown list available above the table. The default is "Topics", but the options include:
Topics: Out-of-the-box topics and topics created by your organisation
Themes: Themes that topics are associated with
Reason for contact: Conversations grouped by reason for contact
Metadata: Metadata relating to your imported contacts (group by a given field)
Managing your columns
In line with the other tabs within the imported contacts page, you can use the manage settings dropdown to change the columns displayed and the order of columns in the insights table.
Click the manage dropdown to show the list of columns within the table. Those with a blue tick are displayed on the page — you can deselect any of these to remove that column from the table. You can change this again later.
You can also drag and drop from the list to re-arrange the order of the columns within the table.
Updates are saved for your view of the table. Other users within your organisation can alter this view to their own individual preferences.
Reasons for contact: The reasons found for a customer contacting you
Topic distribution: Topics found on conversations
xNPS distribution: xNPS score and distribution across conversations represented in the table
Metadata: Select a metadata option to populate the chart
Using the charts
Charts are shown as donut charts by default, but you can click the bar chart icon in the top right to switch to a bar chart view. Use the save image option to download charts as PDFs.
Explaining the table data
Below includes more information on how the data is calculated within the table. These calculations change depending on the grouping. If you're unsure or need a reminder while navigating the page, you can hover over the column header for more information.
When grouped by topic/theme
Column | How this information is calculated | Data type |
Count | The number of eligible conversations associated with this topic. A conversation is "eligible" if its import date is on or after the topic's creation date, and it has been analysed. | Number |
% of Conversations | The percentage of all eligible conversations that include this topic. Calculated as (tagged conversations / total eligible conversations) x 100. | Percentage |
Customer Sentiment | Shows how customers felt during these conversations, on a five-point scale from extremely negative to extremely positive. Each segment in the bar represents the count of conversations for that sentiment. | Colour barometer (red for negative, yellow for neutral and green for positive) |
Agent Sentiment | Shows how the agent's tone was assessed on a five-point scale from extremely negative to extremely positive. Each segment in the bar represents the count of conversations for that sentiment. | Colour barometer (red for negative, yellow for neutral and green for positive) |
% of Repeats | The share of eligible conversations flagged as repeats, based on the xRepeats metric. Calculated as (repeat conversations / total eligible conversations) x 100. | Percentage |
xNPS Score | An xNPS score for these conversations, ranging from -100 to +100. Each conversation is classified as a Promoter, Passive, or Detractor. The xNPS score subtracts the percentage of Detractors from the percentage of Promoters in the topic's eligible conversations. | Number |
xResolution | The share of eligible conversations flagged as Resolved, based on the xResolution metric. Calculated as (Resolved conversations / total eligible conversations) x 100. | Number |
xVulnerability | The share of eligible conversations flagged as containing Vulnerability, based on the xVulnerability metric. Calculated as (vulnerable conversations / total eligible conversations) x 100. | Number |
Avg. handling time | The average handling time across conversations. | Average in days, hours and seconds |
Avg Silence | The average time of all recognised periods of silence. Each conversation's total valid silence is calculated, then averaged across all eligible conversations for this topic. | Average in days, hours and seconds |
Total Silence | The total duration of recognised silence periods in all eligible conversations for this topic. Each conversation's valid silence (over 5s) is summed and the combined total displayed here. | Total in days, hours and seconds |
Avg. Overtalk | The average count of overtalk events per conversation. An overtalk event occurs when both parties speak simultaneously for more than 1 second. | Average in days, hours and seconds |
When grouped by reason for contact there is a slight difference in some of the calculations, as these relate to conversations where a reason for contact was found:
Column | How this information is calculated | Data type |
Count | Number of conversations assigned to this reason for contact. | Number |
% of Conversations | The percentage of all eligible conversations that have been tagged with this reason for contact. Calculated as (tagged conversations / total eligible conversations) x 100. | Percentage |
The remaining columns are calculated in the same manner for the conversation count in the table above. |
Drill down
Drill further into the detail by clicking on key insights of interest.
View by Topic
When grouping the data by topic you can click on insight data to further understand the related conversations and start building up a new filter as you explore the information.
Agent Sentiment
Click on a sentiment colour (red, amber or green) to explore the conversations relating to a given topic where agent sentiment is negative, average or positive. When clicking on the sentiment type you'll be redirected to all these conversations. The filter in the left hand panel will have the topic and sentiment conditions added.Customer Sentiment
In line with the agent sentiment behaviour, you can click on the barometer colours (red, amber or green) to explore the conversations with the given topic and customer sentiment. The imported contacts filters will update with these conditions.% Repeats
Click on the percentage of conversations that are a repeat for the given topic. In the conversations tab you'll find all conversations that have been tagged as a repeat call. To update this to find all calls that are not a repeat, you'll need to change the condition manually from the filter panel.xResolution
Click on the percentage of conversations that were resolved for the associated topic to view all the conversations that were resolved for that topic. In line with other click-through behaviour, the "resolved" and "topic" filters will be populated.xVulnerability
View by theme
These are the themes your topics sit under. You can click on the theme to view the topics related to the theme.
As theme is not an imported contact filter, further drill-downs are not available on the table columns.
You can add other filters directly in the filter panel to narrow down your conversations sample.
View by reason for contact
Reasons for contact are not a filter available in the imported contacts filters, so further drill-downs are not available on the table columns.
You can add other filters directly in the filter panel to narrow down your conversations sample.