Happy New Year!
LivePerson is kicking off 2021 with a number of important updates and exciting new features. Watch this 1 minute video for the upcoming features being released in January.
Intent Manager 2.0
The vision of the Intent-Driven Business is to fundamentally change how brands understand and act on what their consumers want and need - not through proxies like website click patterns - but in consumers’ own intents. Intent Manager is foundational to this vision enabling brands to Discover, Build, Analyze, and Optimize state of the art ML models for intent classification with intuitive self-service tooling.
Intent Manager 2.0 introduces a wide set of new self-service features and functionality that greatly improve the value and utility of the application making it easier than ever for brands to fully manage their Intents on the LivePerson Platform.
Here are all the new features and exciting enhancements coming to Intent Manager.
Topic Clustering for Intent Discovery and Optimization
Optimizing the coverage and quality of intents is a critical component of Intent Management. Intent Manager 2.0 introduces new language clustering visualizations that help brands discover new topics and trends in their conversational data. Brands can combine the power of clustering with Semantic Search and power filters to quickly build new, high-quality intents, which are based on a diverse set of training data that has been extracted from real customer conversations. For more information, click here
A new Model Tester for Intent Builder
It’s important to train and test updated domain models before they are activated. Our new model tester enables brands to more easily see how model changes impact the quality of their intent recognition, as they work to optimize their taxonomy. Brands can use an intuitive interface to create a gold test of phrases, and then map them to their expected intents, just like how a quiz has an answer key.
It’s simple to use:
- Create a test set by adding the sample phrases that you want to ensure that your model is properly detecting.
- Match the sample phrases to their intents. For example, the phrase “I want to cancel my policy” should be mapped to the “cancel policy” intent.
- Once all the phrases are added, click on “Run Test” to generate a score for the current intent model.
Whenever changes are made to the model, you can re-run the test set and review a quality report prior to activating the new model in production.
An Enhanced Intent Trends dashboard with new Customizable Metrics & KPIs
The ability to analyze and track key performance indicators by intent is a key value proposition of the Intent Powered business. Intent Manager 2.0 adds 7 new intent correlated metrics to the Intent Trends dashboard that are aligned to the LivePerson 4E Framework. The release also adds a new configuration panel that enables brands to select which metrics to include in summary cards, charts and graphs, and data tables in the Intent Trends dashboard. For more information, click here
Enhanced Conversation Details Panel for Improved Transcript Analysis
The ability to analyze and explore conversations by intent is greatly improved in Intent Manager 2.0. Brands can use power filters to search for keywords and phrases in both the agent and consumer portions of the transcript. Brands can easily browse a list of conversations that match their filter criteria with associated metadata and KPIS. The transcript view now supports structured content inside a streamline UI.
The Ability to Create and Manage Rule-Based Power Filters
Building on existing advanced filter functionality brands can now save and manage complex filters that enable them to surface and track specific insights on exact topics. Brands can use regular expressions to instantly mine and query up to 13 months of conversation data that matches specific criteria. This functionality is complementary to Machine Learning powered Intents and allows brands to cover analysis use cases that don’t work within the Intent taxonomy.
Added Coverage For Existing Starter Packs
IM 2.0 includes updates to three of core starter packs to improve coverage and accuracy.
- The Telco starter pack now includes 21 intents with typical coverage of 45%
- The Airlines starter pack now includes 13 intents with typical coverage of 74%
- The Insurance starter pack now includes 12 intents with typical coverage of 78%
If you want to create a new starter starter pack model - please contact your account team.
Visual Overhaul and UX improvements
Updated styles, menu structure and usability & accessibility improvements.
Check out the Knowledge Center for more information on Intent Manager 2.0.
LivePerson is excited to launch the Instagram Connector in General Availability.
The integration with the Instagram API allows brands to respond to consumers at scale through the LivePerson Conversational Cloud platform. It enables brands to use automation and rich content in the conversation, and passes context about the consumer’s intent (when available), which can be used for routing, useful info for the human agent, and reporting purposes
What are the entry points?
- Respond to consumers’ direct messages in Instagram (including images and videos file attachment).
- Sending a private message to a consumer who posted a story and tagged the brand.
- Sending a private message to the brand who posts a story about their products and services.
Instagram features include:
- User profile for agent workspace: See the full user profile including name and profile picture of the consumer in the agent workspace.
- Secure form support: Agent can send a secure form to a consumer in a messaging conversation.
- Consumer to agent media share as a file attachment (image and video): Consumers can share images and videos using a file attachment widget in the typing window in native app during the conversation.
- Quick replies from agent to consumer: Agent can set up quick replies.
- CSAT survey: CSAT question will be sent to consumer upon closing the conversation by the agent.
- Post Conversation Survey support: Agent or the custom bot will be able to trigger the Post conversation survey at the end of the conversation.
- Instagram story mention by the consumers: Consumers will be able to mention the brands in their story.
- Instagram brand story reply: Consumers can reply to the story posted by the brands and agents will be able to see the reply.
There will be a number of updates to the Performance Optimizer at the end of the month which will increase user experience and reporting capabilities.
- Additional 4E metrics included within 4E framework: These new metrics have been incorporated into the Performance Optimizer that allow users to further contextualize and optimize their business operation. These metrics include:
- Human First Response Time After Bot (Minutes)
- Human First Responses After Bot Count
- Agent First Response Time Overall (Minutes)
- Agent First Responses Count
- Agent Response Time Overall (Minutes)
Diagnostic Playbooks added to provide tactical optimization plans for each 4E metric.
Resource vs Demand Panel: The Login Hours and Conversation Volume has been replaced by a new panel, "Resource vs. Demand" that displays agent labor hours compared to conversation volume trends to assess whether labor is aligned with conversational volume. The cumulative volume over time has been removed as this did not provide intended value for the user.
Ability to exclude certain 4E metrics within Configuration tab: As some brands are unique and may not leverage specific 4E metrics to analyze their operational performance, the Performance Optimizer will provide the ability to disable these metrics so that they are not displayed within the dashboard or calculated into the 4E composite scores. This provides relevant and accurate 4E performance results based on a brand's operation model.
Ability to start tagging Line Of Business (LOBs) by purpose: Users will be able to to capture the primary and secondary “Line Of Business Purpose(s)”, so that in the future they can associate the purpose of an LOB(s), assess performance and benchmark my LOBs against LOBs with similar purpose within the same vertical.
Opportunities panel redesign: This provides the user the ability to filter by 4E within the Opportunity panel so they can easily focus on specific category metrics. Enhancement also includes color coding 4E metric based on their respective 4E category.
Addition of In App and Web channels to the Channel Panel: This enhancements provides users the ability to see App and Web entry points in the Channel Panel by Source, so that these two entry points are no longer categorized into “other”.
Complete in-depth analysis with data to text diagnostics: The Diagnostic portion of the Performance Optimizer will provide users tactical diagnostics by metric, previously by 4E category, trending data and target deviations are displayed within headers, and the user is guided to additional Conversational Cloud locations to further complete a root cause analysis.
Messaging window enhancements
LivePerson are introducing a number of Messaging window enhancements:
Orient Welcome Message Quick Replies: It will be possible to orient the Welcome Message quick replies into a vertical list, not only a horizontal scroll. This can be configured in the Welcome Message screen within the Engagement Window Studio.
Configure message history in Engagement Window StudioBrands can now configure collapsing message history in the consumer window. Utilizing this option reduces initial window load time and improves consumer experience for consumers with long message histories. We will be slowly changing the default for all brands to YES in the coming months. This setting can be changed at any time in the Engagement Window Studio.
LivePerson are excited to introduce the Bot Performance Dashboard
The new ’Bot Performance Dashboard’ will allow brands to review and analyze the Bot Performance in messaging over time, and measure the impact of agents' annotations on the Bot Performance for Bots designed by Conversation Builder.
The new dashboards will enable brands to track the Bot funnel: Conversations Assigned, Conversations Handled, Intent matched rate and Conversation closed by the Bot. The data will be available by date and hour.
The report will show data at 2 levels:
- Brand level - allowing to view the Bots quality trend over time and and the annotations impact over time
- How many conversations are assigned to Bots? How many are transferred? How many are escalated to other skills?
- What’s the intent match rate? Is it getting better over time?
- What is the annotation rate over time? What is the ROI of the annotations?
Bot level - allowing to analyze and optimize every Bot
- Does the Intent match rate improve when we are investing the time to annotate ?
- What is the trend of unmatched dintents ? Has it decreased over time ?
- Which Bots annotations are submitted ? Does all Bots improve over time when annotations are done by agents?
Updates to Manager Workspace
LivePerson have some important updates to the Manager Workspace
Updates to Manager Workspace
Manager Workspace: Skills level view
Currently, when tracking the performance of multiple skills, managers are required to manually filter the workspace skill by skill, in order to view the data they need. A new skill- level table will be added to the workspace, allowing managers track the performance of multiple skills easily. In addition, the new view will allow users to track additional new metrics as "Actionable" and "Wait time for assignment (50%)".
Manager Workspace: Conversation Intent Filter (Manager Workspace)
For accounts using the Intent Analyzer tool, the intents column in the Conversation Widget will now become filterable. The column which holds the list of all unique intents detected in the conversations will now also enable filtering of specific intents, allowing for easy and quick discovery of conversations on specific topics as well as issue detection.
Please note: In order to see the intent column filter, Intent Analyzer tool must be enabled and configured in the account.
Manager Workspace: Freeze Global Filters
The top pane of the manager workspace will now be "static" and will allow the user to see the global filter when scrolling down. Freezing the headers (making them “sticky”) enables users to constantly see what is impacting their workspace and be able to make changes more easily.
"Online" and "Away" load in load panel of activity summary
Users of the Manager Workspace need to be able to to understand the actual current load on online agents and away agents (human agents only) as a derivative of the total load. These metrics allow managers to heave better visibility into their actual availability and agent’s load. Online Load - The number of active conversations as a percentage of max concurrent conversations of agents in the online state. Away Load - The number of active conversations as a percentage of max concurrent conversations of agents in the Away state.
Updates to Mobile SDK - Hyperlinks
The iOS Mobile SDK will now support rendering links as hyperlinks automatically when adding them in 'hiMessage' and 'Welcome Message'. As part of this feature we will create automatic hyperlinks for URL links, SMS, links, phone number links, and Facetime/Facetime audio links:
- The URL link will open the link in the separate browser
- SMS link will open the Messaging App with the phone number pre-filed in the the "To:"
- Phone number link will open the native iOS action sheet with option to call the number
- Facetime/Facetime audio links opens action sheet with facetime audio, send message and call options.
Developer updates - Functions
LivePerson have updates to Functions including:
- Key/Value Store which provides a simple integrated storage to securely store customer data.
- Functions MTLS support - Functions will be able to enforce an MTLS flow by providing an endpoint and certificate.
Third Party Bot Connectors
This month, LivePerson have many updates to Third Party Bot Connectors:
Transfer to agent: Bots will be able to perform transfer to agent. This allows brands to develop functionality that cherry picks the right agent to transfer and uses to OOTB fallbacks of agent matching system. Conversation Tester: Brands will be able to have mock conversations with their onboarded bot directly from the 3rd-party bot interface. Watson Discovery Search support Brands using Watson: Bots can now also take advantage of Watson Discovery and share the responses of the search back with the consumer via their bot. Dialog flow 'Regions'/'Location' support: Third-party bot platform will now support the regionalization feature of DialogFlow and allow brands to store their bot-related instances outside of the U.S. (default) in other supported zones such as London, Tokyo, Sydney.
Analytics Builder - Permissions updates taking place end of Jan / Feb
We are updating access permissions to the Analytics Builder to ensure that we have an accurate view of active and inactive users. Users who have not accessed the Analytics Builder tool since Sept 1, 2020 will have their permission changed to “off” by default.
For anyone currently using the Analytics Builder, permissions will remain unchanged, but there will be a slight change to the profile name (an additional suffix of _AB will be added). There is no action required by users—this update will happen automatically and no functionality will be affected during this change.
This update will apply to all Conversational Cloud accounts where the Analytics Builder feature is enabled. Contact your LivePerson account team for more information. Click here to learn more
EOL of the Real-Time Dashboard for Messaging
In March 2021, the Real-Time Dashboard for Messaging will be deprecated. The Real-Time Dashboard was developed before the New Manager Workspace existed. The New Manager Workspace is embedded into the core platform, allowing users to manage their operation seamlessly and in real-time which eliminates need to support the Real-Time Dashboard. Note: Chat users will not be affected by this at this time.