Customer Experience News is a weekly newsletter 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.
We don’t have a main story today, simply because my daughter’s last week of school is this week and I wanted to be more present for her while she graduates Kindergarten.
We did launch our second podcast episode of Unresolved.cx that I’d love for you to check out. How do you detect customer disengagement before it shows up as churn? Silent customers slip away, we have the data, AI can automate, but how do we address it?
This is what’s still unresolved for Shmuel Saklad from B&H Photo Video. The full episode can be found on our website along with notes, on Youtube, or Spotify.
Help Scout Expands SLA Tracking to Mid-Conversation Response Times
Help Scout, a helpdesk platform built for SaaS and growing businesses, announced Next Response Time (NRT) goal support within its existing SLA tool. NRT starts a fresh timer each time a customer replies, measuring how long the next support specialist response takes throughout the life of the conversation — not just at the first message or final resolution. NRT compliance is included in SLA reporting.
Operational Impact
Most SLA frameworks have two checkpoints: first response and time to close. What happens in between has been largely invisible. A customer who responds to a specialist’s question and then waits two days for the next reply does not show up as a breach under a first-response-only model. NRT closes that gap. Teams gain compliance data across the full conversation thread, which creates a clearer picture of where tickets stall. The most direct use case is identifying queue patterns — conversations that receive fast first responses but then sit once the exchange is underway. For managers reviewing SLA reports, NRT data adds a layer that first-response metrics alone cannot surface.
Implementation Considerations
NRT does not require changes to existing SLA configurations — teams add it alongside current first response and resolution settings. The practical challenge is calibration. If a team does not know its current mid-conversation response baseline, setting an NRT target requires guesswork. Setting a target too aggressively before understanding actual response patterns is the most common SLA configuration mistake, and it tends to produce compliance data that reflects the target’s unrealism rather than the team’s actual performance. Teams with fluctuating inbound volume, where specialists shift between channels throughout the day, should measure a baseline week before activating NRT goals.
Plain Upgrades AI Replies, Workflow Visibility, and CRM Sync
Plain, a B2B support platform built for developer-first teams, announced three updates across its AI layer and integrations: Ari-powered suggested responses in the composer, a Workflows panel in the Thread Dock, and a native Attio integration that syncs companies and contacts into the support workspace. Each targets a different friction point in the support workflow — drafting replies, tracking automation status, and surfacing account context.
Operational Impact
The Ari update changes the quality bar for assisted drafting. Previously, AI-suggested responses and Ari’s autonomous replies operated at different quality levels — suggestions were lighter, more generic. Now the same iterative search and grounding logic Ari uses when replying autonomously generates the drafts that appear in the composer. For teams not ready to hand resolution to AI fully but wanting to reduce the time support specialists spend on first drafts, this narrows the gap between assisted and automated. Drafts are editable before sending, appear in real time without a refresh, and feedback (thumbs up or down with a free-text note) can flow back into Ari’s evaluation pipeline. The Workflows panel addresses a separate problem: agents managing complex threads often have no way to know which automations have already run. A live status panel showing running, completed, or failed workflows gives specialists the context to continue or escalate without reconstructing what happened. The Attio integration is narrower in scope — it syncs companies from a selected Attio list into Plain as tenants, with people associated with those companies pulled in as customers, refreshing approximately once per hour. Account attributes from Attio display as tenant fields on every thread.
Implementation Considerations
The Ari upgrade is live for workspaces already running Ari — no configuration changes are required. Teams considering Ari for the first time should understand the dependency on knowledge base quality. Ari’s drafts draw from the content it can find; if your documentation is inconsistent or outdated, the suggestions will reflect that. The Workflows panel adds value in proportion to how many automations a team has built — for teams still running mostly manual processes, it surfaces nothing actionable. The Attio integration requires selecting which Attio company list to sync and mapping which attributes to display as tenant fields. That mapping decision carries more weight than the connection itself. Surfacing low-signal fields creates noise for support specialists during conversations where account context should accelerate, not slow down, the response.
Pylon Adds Microsoft Teams Ticketing and Account Data Rollups
Pylon, a B2B helpdesk platform built for Slack and Teams-native support, announced two updates: the ability to create support issues directly from Microsoft Teams chats, and the beta release of Rollups, which aggregates data across related CRM objects into account notebooks. Both target the operational reality of enterprise B2B support, where customer communication is spread across channels and account data is fragmented across systems.
Operational Impact
The Teams update closes a gap that has been a manual workaround for B2B teams running mixed environments. Many enterprise support organizations handle some customers in Slack and others in Microsoft Teams, sometimes within the same account. Without native Teams issue creation, support conversations in Teams have had to be manually converted to tickets or tracked separately from the main queue. Pylon’s update brings those conversations into the same issue lifecycle as Slack threads, which simplifies queue management and gives managers a single view of open work across both channels. The Rollups feature approaches a different problem: account context that should be visible at a glance — renewal ARR, open opportunities, key contacts — has historically required opening a CRM record in a separate tab. Rollups aggregate that data directly in the Pylon account notebook. For support specialists handling escalations where account standing affects how a ticket is prioritized, having that context on screen during the conversation is a practical improvement.
Implementation Considerations
The Microsoft Teams integration requires admin permissions on both the Pylon and Microsoft sides. Teams operating within large enterprise Microsoft environments should expect additional IT approval steps before the connection goes live — this is not a self-serve setup in most organizations. Rollups are in beta, which means the feature is available to test but behavior may change before general availability. The data quality issue applies here in the same way it applies to any CRM aggregation: if field completion in Salesforce or HubSpot is inconsistent, rollup outputs will reflect that inconsistency. Teams syncing CRM data into Pylon should audit field completeness on the accounts they plan to surface in rollups before building any workflow that depends on that data being accurate.
Notion Launches Developer Platform for Agent and Data Integration
Notion, a knowledge management and project management platform, announced the Notion Developer Platform, which lets teams write code to sync external data sources, build custom agent tools, and connect third-party AI agents into a shared Notion workspace. The platform is built around Notion Workers — a hosted runtime for custom code running on Notion’s infrastructure — paired with a new CLI, an External Agents API, and expanded MCP support.
Operational Impact
For support teams that already use Notion as their knowledge base, this release changes what Notion can serve as in the workflow. The database sync capability lets a team pull external data — Zendesk ticket summaries, Salesforce account records — into Notion databases that agents can then query or act on. The External Agents API connects third-party AI agents (Decagon is an existing partner) to Notion as an orchestration layer. A Decagon-resolved ticket could write a summary to a Notion database, which then triggers a Worker to update the account record or route a task to a project. The webhook receiver feature lets any external app trigger Notion directly, closing the one-way limitation that previously made Notion a data destination rather than a decision point. For teams that have been assembling these connections manually using Zapier or Make, Workers offer a code-based alternative that runs on Notion’s own infrastructure without a third-party service in the middle.
Implementation Considerations
This release is designed for teams with engineering capacity. Workers are built and deployed through a command-line interface, which is not a configuration experience accessible to most CX operations teams without developer involvement. CX leaders should be direct with themselves about whether their team has the technical capacity to build and maintain Workers, or whether an integration platform is a more practical fit for the same outcomes. The External Agents API and Agent SDK are both in alpha — not production-ready for most use cases. The pricing model for Workers changes in August 2026, when Notion credits will govern usage after the free beta period ends. Teams exploring the platform should build that cost into any decision made during beta before the feature becomes load-bearing in a workflow.
Fin Adds Voice Note Sending to the WhatsApp Inbox
Fin, the AI customer service platform formerly known as Intercom, announced that support specialists can now record and send voice notes directly from the Fin inbox when replying to WhatsApp conversations. The customer receives the message as a native WhatsApp voice note, and each recording is automatically transcribed for search, export, and AI context.
Operational Impact
WhatsApp voice notes are native to how many customers already communicate on the platform, particularly in markets where WhatsApp is the primary support channel. For teams managing high WhatsApp volumes, voice notes open a response format that communicates differently than text. Technical walkthroughs and service moments that require nuance — explaining a process step by step, or responding to a frustrated customer — can be clearer in audio than in a written message. The auto-transcription means the exchange stays searchable and auditable, which matters for quality review. Teams using Fin’s AI context should also benefit from transcribed voice content being available as part of the conversation record, since AI-generated suggestions in subsequent turns can draw from what was said in the voice note rather than treating it as a gap in the thread.
Implementation Considerations
Voice notes are relevant only for teams where WhatsApp is an active support channel. For teams not using WhatsApp, there is nothing to evaluate here. For those that do, the practical question is whether the work environment supports audio recording — open-plan offices create friction that makes this feature less usable than it looks on paper. Specialists also need clear guidance on when voice is the right response format: recording a clear, useful voice note takes longer than typing a short reply, and using it indiscriminately can add queue time rather than reduce it. Transcription accuracy depends on recording quality and language, and teams in multilingual or high-noise environments should test the transcription output before relying on it for AI context or compliance review.
Linear Brings AI Codebase Access to Support and Product Teams
Linear, a project management platform built for software teams, announced Code Intelligence, which gives Linear Agent controlled access to a connected GitHub codebase and makes that context available to everyone in the workspace. The feature is in public beta for Business and Enterprise plans and free to use during the beta period.
Operational Impact
The most immediate impact for support teams is the ability to answer technical customer questions with more grounding. When a customer files a ticket about unexpected product behavior, a support specialist can query Linear Agent about how the relevant feature is implemented without waiting on an engineer to explain it. That removes a step in the escalation chain for questions that are answerable from the codebase without requiring a code fix. For product managers, the same access means specs can be written with direct reference to what the code actually does rather than relying on documentation that may not reflect the current state. Code Intelligence addresses a specific gap in cross-functional teams: engineering knowledge has been largely inaccessible to non-engineers without a direct request, and those requests carry a cost in time and context-switching on both sides.
Implementation Considerations
Setup requires the GitHub integration to be installed and code access enabled by a workspace admin. Admins can scope access to members with existing GitHub permissions or open it to the full workspace — that decision carries real weight for organizations with proprietary or sensitive code. Support specialists using Code Intelligence to respond to customer questions should treat AI-generated codebase explanations as a starting point rather than a definitive answer. Code behavior is not always evident from a static read of the repository, and an incorrect explanation communicated to a customer can deepen the confusion rather than resolve it. Teams building any workflow where Code Intelligence output goes directly to customers should include a review step before that becomes a pattern.
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.
