ServisBOT multiple ways to measure activity and success for your bots and Virtual Assistants.
By default, ServisBOT collects a set of messenger, conversational, and system metrics which are accessible in different parts of the platform. Some configuration may be required to view or explore this data, some of these may only be available when using the enhanced messenger.
Custom KPIs can also be tracked by using goals and usecases
These metrics track engagement on the individual host pages where bots or VAs are made available. Generally these are the interactions before an end-user interacts with the bot.
Some examples include
These metrics are only available for enhanced messenger customers.
Total number of times a web page is loaded where the messenger or messenger call-to-action (roundel) is visible
Total number of unique devices that visited a page where the messenger or messenger call-to-action (roundel) is visible
Conversations started is available in two sections of the servisBOT dashboard but tracked in the same way (Messenger Engagemant and Chat engagement)
Total number of times a user initiated a conversation with the messenger through a “conversation start action:
autoStart.Ratio of unique impressions that resulted in an started conversation. If a stronger percentage of users who come to your site decide to start a conversation, then your messenger engagment strategy is working well.
These metrics track activity after user initiates a conversation. These metrics will give you an overal understanding of your user’s demand for a chatbot experience.
Total unique trackable users with conversations that had at least one active session started per calendar day. Users can come back to your site and engage with the bot more than once, or never come back at all. This metric provides you with insight into general usage, per calendar day. This is also how you will be billed.
customerReference that maps to any user ID you have available for users so that our kown DAU is driven off of your known user state.CustomerReference is provided on init of messenger, one will be generated when the messenger is loaded for thef first time for a user.Count of conversations where the user sent at least one message or interaction into the conversation. Conversations can only be tracked as engaged once until they expired.
Ratio of started conversations to engaged conversations. Not every conversation that is started by a user results in the user sending your chatbot a message. This rate will measure how well your welcome experience is designed.
Average rating of all conversations where a rating was provided. In order to track this rating a Feedback node needs to be used somewhere in the conversation
Breakdown of rating values and number of times that rating was selected for a conversation. In order to track this rating a Feedback node needs to be used somewhere in the conversation
The number of messages or interactions sent by end-users. Conversations with lots of user inputs is a sign of good engagement.
Breakdown of interaction types and how often they happen
Total Intents triggered in this experience for the selected date.
Ranked list of most popular intents across this bot.
Sessions track a period of activity that a user has with a conversation. Session length is configurable per experience, but is recommended to be 20-30 mins, or at least less than a conversation expiration. Conversations with multiple session are another way of understanding that users are engaging with a bot, beyond just starting conversations for the first time.
Total count of sessions started. A session defined as a group of user interactions within a messenger that takes place within a given (configurable) time frame. A single conversation can have multiple sessions. Each conversation will have at least one session. Session periods can be configured between experiences. It is recommended to be 20-30 mins, or at least less than a conversation expiration.
Total count of sessions started outside of a dashboard’s configured business hours. Business Hour Sessions
Average number of user inputs that occured within a started session. This provides a deeper drilldown into conversational engagement.
For customers using channel shift, ServisBOT automatically collects a set of metrics related to the end-users experience.
Count of instances where a call was transfered to the ServisBOT channel shift IVR.
Count of mobile or non-mobile calls into the ServisBOT channel shift IVR
Result of SMS or Email send requests
Total times each unique URL generated by channel shift is rendered. Messengers rendering a page preview will track as clicked links.
Tracking interactions and data related to any point within a conversation can be done using ServisBOT’s Goals which records that a particular KPI was hit in a conversation. Conversation_Start and Conversation_Engage are tracked as goals by default.
Collections and groups of goals can be made using the category field in a goal. This allows tracking of interactions within specific conversion funnels of a conversational experience. Metrics grouped this way can be added to the ServisBOT dashboard using additional configuration.