Conversations Page Overview

On the left-hand side of the Conversations page, you'll find a list of conversations, ranked from the most recent to the oldest within the selected date range. Simply scroll down the list to view more conversations.

❗️

By default, the conversations list displays conversations that occurred within the last 48 hours. You can adjust the date range using the available filters.

Conversations Parameters

User Name

The user name is displayed if available in the user's profile on a specific (e.g., Facebook Messenger, WhatsApp, WeChat). This information is shown only when the user is authenticated on that channel.

User name displayed from the user's profile on Facebook Messenger.

If the user's name is unavailable, such as when using the ViaSay Chatbot (even if logged in on your website), the user is labeled as “Anonymous User”.

Anonymous user on the web widget.

User Channel

An icon on the left side of each conversation in the list indicates the channel where the conversation occurred.

Agent Handover

If an agent handover protocol is enabled for your bot, conversations transferred to an agent are tagged accordingly. This tag appears only if the user was successfully routed and the agent joined the conversation.

Conversations handed over to an agent.

Bot-Initiated Conversation

A bot-initiated conversation starts with a push message, such as a WhatsApp push or Scheduler push. These conversations are identified by a green speaker tag.

Issues Created

When an issue is created from the Conversations page, a tag appears next to the conversation in the list, indicating that an issue has been logged.

Tag indicating that two issues were created in this conversation.

Conversation Score

ViaSay Chatbot automatically rates the bot's performance during a conversation on a scale of 1 to 10. Each bot message is assigned a score based on the following criteria:

• Score A: Message triggered by a quick reply.
• Score B: Message responding to a typed message (lower than Score A).
• Score C: Message responding to an out-of-scope request (lower than Score B).

The conversation score is the average of these individual message scores.

Conversation Date

The date of the first user message in the conversation is displayed on the right side of each conversation in the list.

Number of Messages

The total number of messages exchanged between the bot and the user is also shown on the right side of each conversation in the list.

Conversations Filters

The filters at the top of the conversations list allow you to narrow down the conversations by three criteria:

• User Channel Configuration
• Conversation
• Bot Architecture

Total Conversations

The total number of conversations fetched is displayed at the top of the list. This number updates whenever a filter is applied.

Approximately 2,000 conversations match the filters applied here.

Click the Filters button to open the Filters panel, select the desired filters, and click Apply to update the conversations list.

Filters panel.

Filter by User Name (Above the Filter Panel)

The search bar above the Filters panel allows you to filter conversations by user name when available.

User Channel Configuration

By Platform: Select the platform where a conversation occurred (e.g., Facebook Messenger, Widget).

By Bot Channel: Choose the specific User Channel where the conversation took place. This is useful when the bot is displayed on multiple pages with different Widget Configurations for each page.

By Language: Filter conversations by the language used.

Conversation

By Date Range: Display conversations that occurred within a specific date range.

By Conversation Score: Select conversations based on a range of conversation scores. This is particularly useful for reviewing poorly rated conversations to identify areas for improvement.

By Live Agent Handover: Filter conversations based on whether they were handled solely by the bot or involved a live agent handover. By default, both are displayed.

By Initiation: Filter conversations based on whether they were user-initiated or bot-initiated.

Bot Architecture

By Use Case Triggered: Filter conversations based on a specific Use Case triggered by user messages.

By Intents Triggered: Select conversations involving a specific <glossary:Intent> triggered by user messages.

By Bot Actions Triggered: Choose conversations involving a specific Bot Action triggered by user messages.

❗️

You cannot filter based on deleted intents, use cases, or bot actions.

Conversation Details

On the right-hand side of the Conversations page, all interactions between the bot and the user in the selected conversation are displayed.

Example of conversation details displayed on the right-hand side of the Conversations page.

❗️

A user can be associated with multiple conversations. For example, if a user leaves the chat window open and is inactive for 15 minutes, the conversation will end. If the user starts interacting with the bot again, a new conversation will be triggered. This explains why a user might be linked to several conversations.

• The User Channel icon is located in the top left-hand corner.
• The User Name is displayed next to the User Channel icon when available.
• In the top-right corner, a message search bar allows you to search for specific words within the selected conversation's messages.
• The date of the first user message in the conversation is displayed at the beginning of the conversation. The time of each message is shown beneath it.

❗️

The time displayed is in the UTC timezone.

User Messages

User messages are displayed on the left side of the conversation. Additional indications are provided if a button or Quick Reply is clicked. If no indication is provided, the message was typed by the user.

Example of a user message sent via Quick Reply.

Clicked Links

You can easily see when a user clicks on a web link provided by the bot. This is a good indicator of the conversation's success. You can click on the link to access the webpage the user was redirected to.

Example indicating which buttons the user clicked.

Bot Answers and Push Messages

On the right side of the conversation, you'll find the bot's answers and push messages.

Agent Messages

When a live agent successfully takes over a conversation, their messages are displayed in the conversation thread.

Example of agent messages.

Actions on Messages

When hovering over a bot answer, three icons appear, allowing you to perform several actions:

Copy Direct Link

To copy a direct link to a specific message:

1. Hover over the message and click on the three dots icon that appears.
2. Select the Copy Direct Link option.
   This action allows you to copy the link to the conversation and directly access the selected message.

Copy Direct Link button.

Daily Conversations Extract

This feature allows you to export your bot's conversations on a daily basis in a CSV file. If you would like to activate this feature, please reach out to us.

Each day at 3:00 AM UTC, ViaSay will export a CSV file containing the previous day’s conversations and make it available on an AWS S3 bucket. This S3 bucket can be hosted either on ViaSay's AWS account or on your AWS account, based on your preference.

These exports contain anonymized data (e.g., without credit card numbers, emails, or phone numbers). If the extracted files are hosted on ViaSay's AWS account, they are automatically deleted after 40 days.

Hosting

The exports can be stored in two different ways:

1. S3 Bucket Hosted on ViaSay AWS Account

ViaSay Chatbot will provide you with credentials to access the bucket. ViaSay customers can use any of the methods provided by AWS S3 to retrieve the files.

2. S3 Bucket Hosted on the Customer's AWS Account

ViaSay Chatbot can also use a shared AWS S3 bucket. The complete instructions for creating an AWS S3 bucket with cross-account access can be found here.

Out of the several setup options, ViaSay Chatbot recommends the “Resource-based policies and AWS Identity and Access Management (IAM) policies” method.

Full read and write access is preferable at the beginning of the project as it eases integration tests. If full access is not possible, ViaSay should at least be provided PUT, LIST, and DELETE rights.

Export File

File Name

Each day, ViaSay Chatbot uploads one conversation file for each bot (if the bot has had at least one conversation with a user) with the following naming pattern:

<date>_<bot_name>_daily_conversations_export.csv

For example, the file containing conversations from January 21st, 2022, for a bot named "Example Bot" would be named 20220121_example_bot_daily_conversations_export.csv.

All exports follow the same format. The CSV file contains columns as described in the next section. The column separator is a comma, and the string delimiter is a double quote.

Content and Column Names

Before we dive into the column names, let’s define the key terms “end-user” and “conversation”:

• End-user: A user interacting with a bot.
• Conversation: A series of messages exchanged by an end-user, a bot, and, potentially, a customer service agent. A conversation ends if an end-user does not send a message within 15 minutes. More precisely, a conversation is a group of messages with at most a 15-minute gap between two messages.

When handling a new incoming message, there are three scenarios:

1. If this is the first message, a new conversation is created.
2. If the last message was sent more than 15 minutes ago, a new conversation is created.
3. If the last message was sent less than 15 minutes ago, the message is added to the current conversation.

The CSV file contains one row per message. The columns of the file include: