Customer Experience News is a weekly newsletter and video about the most important news and discussions for Customer Experience and Customer Support Leaders.
This is all the weekly news you need in around 5 minutes.
October is the busiest month of the year for conferences across Customer Experience, Customer Success, Customer Support, and Contact Centers. Vendors and leaders are traveling to attend and present, and that makes it easy to miss product updates. This issue collects the updates worth catching up on, with one section for each tool.
Today’s headlines are from:
Fin (Intercom) Adds Staffing Forecasts and Macro Analytics
Intercom, a customer service platform, announced Workforce Management, along with a Usage tab for Macros and the ability to follow conversations and tickets. Workforce Management covers volume forecasting, reusable schedules, a coverage view that compares required and scheduled teammates, and a Team Performance review of how a day went. The Macros Usage tab shows how often each macro is used, which teammates and teams use it, and in which conversations, for any time range. Following lets a teammate stay informed on a conversation or ticket without being assigned to it.
Staffing plans that live in spreadsheets move into the system where the queue runs, so a support lead can compare required coverage with the schedule without exporting data. The Usage tab sorts macros by usage count or last-used date and supports bulk delete and CSV export, which gives a team a practical way to retire macros nobody sends and to see which ones carry the most traffic. Followers receive the same desktop, mobile, and email notifications as the assignee, and a Followed by filter in Search and Views builds a list of everything a teammate follows. A manager can track an escalation or a key account without taking ownership of it.
Workforce Management requires the Pro add-on, and Beta users need it to keep their existing setup. The announcement does not describe the forecasting method or its inputs, so a team should compare the forecast with its own historical volume before building schedules around it. Reporting access grants default access to Workforce Management, and admins can change access per teammate. Each conversation can have up to 100 followers, and followers carry over when conversations are merged. Because followers receive assignee-level notifications, teams may want a simple norm for who follows what. The Macros and Followers announcements list no plan requirements.
Read more: Workforce Management →
Gorgias AI Agent Reaches Socials, Adds Skill Trends
Gorgias, an ecommerce customer support platform, announced that AI Agent on Socials is generally available for Instagram DMs and Facebook Messenger, plus WhatsApp. AI Agent uses the knowledge, skills, guidance, actions, and settings already configured for its other channels, and teams can set a separate tone of voice for social. Shoppers can share an order email or number and get help inside the thread, and the Shopping Assistant recommends products with add-to-cart in DMs. Conversations that need a person go to the team in the same thread under existing handover rules. Gorgias also added a trend chart to each skill’s Performance tab that shows handovers and CSAT over time, with markers for each published version of the skill.
A team that already maintains AI Agent for other channels can extend that configuration to social without a second build, and Channel Analytics filters AI Agent performance by social channel. The trend chart answers a question that current numbers alone could not: whether a recent edit to a skill helped or hurt. It shows the last 28 days by default, accepts a custom range, and lets a team match each version marker to the movement in handovers and CSAT that followed.
Setup requires an active Facebook and Instagram integration, a WhatsApp integration, or both. In AI Agent, go to Deploy and then Socials. Instagram DMs and Facebook Messenger turn on together, and WhatsApp can be enabled separately. Beta users need no action. Because social conversations share handover rules with other channels, teams should confirm those rules fit the pace of DMs before turning the channels on. The trend chart covers two metrics, so a team judging a skill change should also watch volume and topic mix over the same period. The announcements do not state plan differences or limitations.
Front Expands Agent Tools and Account Context
Front, a customer communication platform, announced a monthly update centered on coordination between people and Agents. Teams can coordinate customer work across Support Specialists and Agents with shared context, and guided handoffs carry that context along. New evaluation and testing tools let teams assess Agent performance and try out how Agents handle customer work before deployment. Accounts link account context to customer work, AI summaries condense customer history and next steps, and Copilot routines deliver meeting briefs, follow-ups, and churn-risk alerts to the inbox. Reporting gains expanded conversation exports, task metrics by teammate, inbox, and tag, more analytics metrics in the reports API, and a token-level breakdown of API usage.
Testing Agents before they reach customers gives a team a place to catch problems early, and the evaluation tools support ongoing tuning after launch. Account context in the inbox helps a Support Specialist see which customer relationship a conversation belongs to without switching tools. The token-level API breakdown helps teams investigate traffic spikes or throttling. Front also added a Modem integration that groups conversations into feedback topics, and rules that require tag lists, custom field groups, or both. When a field group is required, every field in it must be filled before a conversation can be resolved, which raises the quality of data that reporting depends on.
Guided handoffs are only as good as the context captured upstream. Required tag and field rules add steps for Support Specialists at resolution, so test them on one team and watch resolution times before extending them. Account features depend on how cleanly accounts are set up in Front today.
Plain Tightens Sidekick Permissions and Updates Ari
Plain, a B2B customer support platform, announced a broad set of changes to its Agents and workspace. Ari, its customer Agent, now sends at most one handoff message per conversation, continues investigating from the answers customers give to its clarifying questions, and runs on updated models for faster responses. Sidekick, its internal Agent, now limits tool calls to what both its API key and the requesting person can access, and it can read GitHub pull requests and issues with read-only access. A View gaps button in the Ari Overview lists knowledge gaps with age and assignee, and Sidekick follows a saved article format when drafting articles to fill them. Snoozed and Done queues load faster, and a mobile thread view supports replies, notes, and actions from a phone.
The knowledge gaps table gives a documentation owner a worklist with names attached, and Sidekick drafts the articles that close each gap. Hover previews show the customer, account, channel, summary, labels, assignees, and any breaching SLA without opening a thread. A new Threads re-opened metric shows how often threads reopen after being marked Done, and an AI translation option lets teams read messages in other languages inside a thread. Plain MCP can now find, create, and update tasks, read knowledge gaps, and request source reindexing.
Enabling GitHub reading requires approving the updated read permissions for the Plain GitHub App, and access is limited to repositories selected in the integration. Automatic workflows still use the API key’s access, so teams should review the scope of those keys. Threads that Ari handled in Test Mode can be switched to Live, and the Not assigned view finds threads with only Test Mode Ari assigned. Message translation is off by default and admins enable it in Settings under Translation. Lightweight GraphQL queries now return up to 1,000 rows when they select only supported ID and update-time fields, while other queries stay at 100, so integrations that use pagination or change-since filters need an update.
Pylon Launches CSM and CS Leader Overview
Pylon, a B2B support platform, announced the CSM and CS Leader Overview, a customer success home page under Customers, then Overview. It combines a Daily Overview of meetings, tasks, and issues waiting on the user with Signals that flag churn risk, renewals, expansion, and sentiment. Charts show revenue and account health, and the CS Leader template adds GRR and NRR charts. A configurable AI command bar sits on the page, and admins can use Overview Boards to give each team its own set of views. Everyone starts with the CSM Home view.
Customer success managers start the day from one page that lists meetings, tasks, and the accounts that need attention, and leaders get revenue and health charts in the same place. Signals put churn and expansion flags next to the work due today, so a manager can decide which account to contact first. Pylon shipped related changes in the same release: approval cards that pause AI agent tools set to Needs approval, a Request access option when a tool is blocked, and support Agents that can use private GitHub repositories as documentation sources.
Meetings in the Daily Overview require a connected Calendar integration. The documentation does not say where Signal logic or revenue figures come from, so compare Signals with account histories the team already knows before relying on them for renewal decisions. The CS Leader template may not appear in every workspace, and team-specific views depend on admin setup in Overview Boards. Teams will also want to decide who responds to Agent tool approval requests and how quickly.
Featurebase Adds Workload Management for Queues
Featurebase, a customer feedback and support platform, announced workload management, a set of tools for conversation assignment as queues grow. Balanced assignment routes conversations to available teammates with the fewest open conversations, within teammate and inbox limits. Round robin assignment rotates new conversations evenly. Queue prioritization orders work by urgency, SLA deadlines, or customer wait time. Teams can add backup teammates or use Workflows to send new conversations to another team at capacity, and a live dashboard shows waiting conversations, availability, and team capacity.
Assignment is a daily source of manual work for team leads, and these controls move it into rules. Away mode can be set manually or triggered automatically after inactivity, with an option to reassign a conversation when a customer replies while the teammate is away. The live dashboard gives a lead one view of who is available and where the queue is building.
Balanced assignment and advanced workload controls are available on the Enterprise plan, and the announcement does not say which plan includes the other features. It also does not describe how teammate and inbox limits are configured, so the help center overview is the next stop. Teams should decide how SLA-based prioritization interacts with balanced assignment before turning both on, and should watch reassignments from away mode closely during the first weeks.
Guru Knowledge Agents Can Now Update Answers
Guru, a knowledge management platform, announced that skills-based Knowledge Agents can now update answers, bringing the Edit answer capability of legacy agents to the newer type. In Operator Mode, a user refines an answer through conversation with the Agent and then tells it to update the answer. Guru saves the new text and sources and notifies the person who originally asked the question.
Corrections reach the original asker, which closes the loop on a wrong or outdated answer. Teams that stayed on legacy agents because they needed answer editing now have a reason to move to skills-based agents. A knowledge owner can fix an answer in the same conversation where the problem surfaced.
The feature requires the Manage questions permission, a skills-based Knowledge Agent, and Operator Mode. Source boosting is the one legacy Edit answer capability that is not included, so teams that rely on it should keep that gap in mind. Guru suggests testing on one recent answer that needs correcting and confirming that the original asker is notified before working through a larger backlog.
Notion Adds Column Permissions and Agent Connections
Notion, a connected workspace platform, announced version 3.8 with changes aimed at Agents and governance. Any compatible agent can connect through MCP to read workspace context and update docs and task trackers, and admins can see who connected each AI tool and when it was last used, then revoke one person’s access without affecting other connections. Notion Agents can connect to more than 150 tools, including HubSpot, Stripe, and Cloudflare. A Skills API syncs skills written in Notion to tools such as ChatGPT and Claude. Database owners can now set column permissions that control which properties each person or agent can view or edit, and synced databases for GitHub, Jira, and Salesforce were rebuilt for speed and reliability with live sync health.
Column permissions let a support Agent use account context while restricted customer fields stay hidden, which matters for teams that keep customer data and internal notes in the same database. Rebuilt synced databases work like native databases with custom properties, views, and automations, so engineering escalations from Jira or GitHub can sit beside support data. Upgraded version history shows highlighted edits and who made each change, with the option to restore earlier versions. Microsoft Teams notifications bring mentions, comments, and access requests into Teams.
Admins can limit connections to tools verified by Notion or to specific tools they choose, which makes an early audit of connected tools worthwhile. Column permissions require a decision about which properties are sensitive before they are applied, and that classification is work for the team to do. The release notes do not state plan availability for each feature. Skills synced to other tools should be tested after each change, since one edit reaches every tool that loads the skill.
GitBook Agent Answers Questions About Docs Analytics
GitBook, a documentation platform, announced that GitBook Agent can answer questions about a site’s analytics data. A user clicks Ask GitBook Agent in site analytics, asks a question, and receives an answer along with charts, including charts that no built-in report covers.
Documentation owners can ask for a custom view of reader behavior and get a chart in the conversation, without building a report by hand. For support teams that measure documentation as part of deflection, this shortens the path from a question about content performance to evidence.
The feature is available on the Ultimate and Enterprise plans and requires access to site analytics, where the button is located. The entry lists no other requirements or limitations. Teams should check the Agent’s charts against the built-in reports for a few known questions before using them in reviews.
Linear Loops Gain Slack and GitHub Triggers
Linear, a product development platform, announced new capabilities for Loops, its automated workflows that act on product work. Events in Slack or GitHub can now start a loop automatically. Examples include answering product questions in a Slack channel and drafting a QA plan when a GitHub pull request receives the ready for testing label. Teams can save a loop as a template for their workspace, and each loop can be set to a reasoning level of Auto, Low, Medium, or High to balance speed, cost, and task complexity.
Support teams that escalate to engineering work with Linear on the receiving end. A Slack-triggered loop can answer routine product questions in a channel before they reach an engineer, and shared templates let one team publish a bug-triage loop that other teams adapt. Reasoning levels let routine work run quickly and cheaply while harder tasks get deeper reasoning.
The announcement does not mention plans, pricing, prerequisites, setup steps for the Slack and GitHub triggers, or limitations. The Loops documentation is the place to confirm these. Teams should test a loop in a single channel first and review its answers for accuracy, since a loop that answers product questions speaks for the team in that channel.
Now this is:
Strategic Support for CX Leaders.
You’ve got ambitious Support targets and new metrics but you’re not sure what to prioritize first.
The list is long, the queue is getting longer, and you don’t have time to step back and think about CX strategically.
What if you could pressure-test your thinking with someone who’s spent 20 years building Customer Support operations?
No pitch, just a conversation.
