How Pylon uses Pylon

Purpose: Many customers have asked us how we deal with different situations internally, or how we have set up Y feature internally. Here is the much requested deep-dive on How Pylon uses Pylon, covering the highlights of how we handle customer support at Pylon, in Pylon

This article will answer the following Qs for our internal org:

• What customer issues come into our Pylon system?

• Which team works on what customer issues?

• How does an issue route to the right Pylon team?

• After team assignment, how does the issue get assigned to the right team member?

• What are best practices for working Issues in Pylon?

• How do we loop in other team members not in Pylon?

• How do we continuously improve our setup?

What customer issues come into our Pylon system?

We have 1,000+ shared Slack channels with customers and prospects, not to mention all the other ways our customers get in touch. The only way to stay on top of everything is by bringing all these communications channels into Pylon:

• Customer Slack channels, e.g. #pylon-hightouch. At this point, most team members are not in these channels, and interact with customers solely through the Pylon web app

• Email, with separate domains each forwarding into Pylon for our respective teams (e.g. support@, success@, billing@)

• Chat Widget. In our customer's Pylon web app, you can reach out to our support team via the widget in the bottom right

• Customer Portal, where customers can submit Ticket Forms

Notable exceptions:

• Inbound emails to a team member's email: In this case, we will cc the appropriate Pylon email (e.g. support@, billing@) when responding to the customer and then the thread will track in Pylon

• Slack DMs: This is by design as there are cases where the conversation is irrelevant or distracting to the entire shared Slack channel

Which team works on what customer issues?

Here is our team structure at Pylon

Each team that primarily uses Pylon is setup as a Pylon Team to enable routing

How customer issues route to the right Pylon team?

Our previous routing scheme (inactive as of mid-Feb 2025)

• All inbound issues from a Prospect or Trial Customer are automatically assigned to our Sales team [Trigger]

• All other inbound issues are automatically assigned to our Support team [Trigger]

• Support team member reviews the issue and manually assigns it to the appropriate team if needed, or works the ticket themselves [Manual assignment]

• The recipient team sees the Issue in their team-specific View [Views]

Current routing scheme

• All inbound issues from a Prospect or Trial Customer are automatically assigned to our Sales team [Trigger]

• All other inbound issues are automatically assigned to our Support team [Trigger]

• New step: When Issues are first assigned the Support team, the assignee is first set to the Router AI Agent. Our Router Agent uses 3 separate Runbooks to route Issues to Customer Success, Migration, and Billing teams, respectively. If the Router Agent does not route to any of these three teams, then Router AI Agent unassigns itself and remains with the Support team to resolve [AI Agent, Runbooks]

• All teams view Issues in their team-specific View [Views]

After team assignment, how does work get to the right team member?

Case #1: Team is pooled (Support, Migrations, Billing, Legal, Security)

Pooled teams are teams where all team members work on everything assigned to that Team. E.g. our Support team members monitor and work all issues assigned to the Support Team

Case #2: Team has well-defined scope for each team member (Customer Success, Sales)

For Customer Success and Sales, each team member has a set of customers / prospects they work on

• Sales: We sync Account Owners from Salesforce. Each Sales team member manages a View of Issues where Sales Team is assigned and Account Owner is themselves [CRM sync]

• Customer Success: We also sync in Customer Tier from Salesforce. Each Customer Success team member creates a View of Issues where Customer Success team is assigned and customer is in specific Tier [CRM sync]

What are best practices for working Issues in Pylon?

Team members work out of their own View and responds to Issues 1 by 1. Here's how we make their lives easier:

• Priority level: Priority is set automatically by AI based on the Issue, and then Issues are ordered on the Kanban board from high priority to low priority. Our team members can work efficiently from top to bottom in the New and On You columns

• Custom Fields: We sync in Salesforce fields that give context about each customer (e.g. what they bought, their renewal date, TCV). Use custom fields rather than tags for structured information that you'll use on every issue thread - custom fields can be made required and offer better reporting capabilities, while tags are better for ad-hoc categorization.

• Macros: Templates for most common responses. E.g. for the Customer Success team, we have macros to respond to meeting requests with a Calendly link and macros for questions about different AI bundles

• Shortcuts: For quickly navigating and responding to issues

• Knowledge base sidebar to quickly copy links in for relevant articles

• AI features:

  • AI copilot that drafts a reply to the customer and cites sources

  • Sidebar copilot for generating a summary for quickly getting up to speed on a long customer thread, or asking an Issue-specific question

  • AI-supported custom fields: Automatically categorize issues using AI by enabling AI support on custom fields. Note that there's a limit of 2 AI-supported custom fields on lower tiers, with unlimited access available on the Enterprise tier. You can check which fields have AI enabled in your custom fields settings. For best results, provide natural language descriptions with relevant keywords rather than just lists of terms.

How do we loop in other team members not in Pylon?

Primarily this is to loop in Product, Engineering, or Design team members

Method #1: Internal threads

For questions to other members of Pylon. Here are the most common internal Slack channels we create internal threads to:

Method #2: Linear integration

For feature requests, we create a Linear ticket in Pylon. These tickets are then on the Product team to review and include on the Product roadmap

How do we continuously improve our setup?

We are always aiming to better support our customers, by being:

• More self-serve

• Faster responses

• Better responses

• More proactive

To improve, we must first understand how we are doing. Daily and weekly, we review two key pages in Pylon web app:

• Analytics dashboards for getting an overall view and reviewing specific metrics to track how we're doing (time to first response, time to response, time to resolution, CSAT, number of touches)

• Knowledge gaps: While working individual tickets, team members mark questions as needing articles. Our Support team meets weekly to review the Knowledge Gaps page for all these marked questions, and churns out knowledge base articles. Our AI Article Generation feature can help here!

• Common topics: Pylon uses AI to group common questions and then ranks them. This is all questions your customers have ever asked. Super powerful, because we can then tackle these top to bottom to determine what is the appropriate action to take. For example:

  • Customers confused by a particular tooltip in the product? File a Linear ticket

  • Customers are getting stuck about a particular step in the Teams integration? Improve existing Teams integration documentation

  • Customers are asking about key considerations when migrating from Zendesk to Pylon? Create sales and migration collateral

One final note: We are pushing for our support to become even more proactive - so we get to you, before you know you need us:

• Broadcasts: Used to send messages to customers proactively about any issues affecting multiple customers, or for product updates

• Ask AI: Used to ask questions summarizing across our issue volume. We have used it proactive to identify feature requests and prepare for customer conversations. Here's our prompts respectively:

  • What are the most common feature requests related to our Customer Portal feature?

  • Can you summarize our issues recently with Customer X?

• Account management (Alpha): This will enable customer-facing teams to have a holistic, 360-degree understanding of any of their customers, to power proactive workflows like reachouts based on low or high sentiment

More examples of how Pylon uses Pylon

We share more of our workflows and how we use certain Pylon features on LinkedIn. You can learn more in these posts:

• How we escalate bugs to engineering

• How we prep for customer success calls

• How we handle incidents

• How we identify seat expansion opportunities

• How we escalate product feedback

• How product researches new features

• How we identify upsell opportunities

• How we gather evidence for product ideas

• How we do leadership reporting for support

• How we do leadership reporting for customer success

• How we keep our knowledge base updated

• How CS shares key account context with support

• How we hand accounts off from sales to CS

• How we personalize onboarding

• How the GTM team learns about our customers