Help center Getting started Start your day and keep support work moving Updated Wednesday, October 28, 2026 · Marc Staine A reliable support routine helps you spend less time rediscovering context and more time moving customer work forward. Use this guide to learn where to look, what to check before replying, and what to leave behind when someone else takes over. In B2B support, the customer relationship often spans several contacts, products, and teams. At a glanceFor: new agents, experienced agents, and team leads onboarding colleagues. Applies to: Supportbench Core.Start here: Cases and Queues in the agent navigation Learn the working vocabulary · Start with commitments and waiting work · Build context before responding · Make the next useful update · Finish the shift with a clear handoff · A supervisor’s onboarding check · Your first ten minutes in the agent workspace · Leave a handoff another agent can act on · When your workspace looks different Learn the working vocabulary TermWhat it means in daily workCompany and contactThe organization you support and the individual involved in a request. Verify both before communicating.CaseThe tracked request or investigation, with state, ownership, activities, and related context.ActivityA reply, note, or other recorded interaction. Its type and audience matter.Agent and queueAn individual responsibility or a team’s pool of work. Check which currently owns the next action.ViewA reusable selection and presentation of cases. A view does not assign or update those cases.Knowledge articleReusable guidance that can help resolve a case or support self-service, subject to its audience and publication state. Start with commitments and waiting work Open Cases. Check the Agent, Queue, and View selectors before interpreting the list. Start with your assigned work and the queues you cover.Use your team’s agreed view for time-sensitive commitments and work awaiting a response. Confirm active filters and the reporting time context.Open the cases that need attention. Read the latest customer message and internal handoff before deciding what to do.Check notifications, mentions, and tasks that apply to your work. Reading a mention does not transfer case ownership or send a customer reply.Agree coverage with your lead if the workload or a commitment is at risk. Use the escalation process when a decision or specialist response is needed. Build context before responding Check the company and contact, current state, priority, division, issue, owner or queue, and service commitments. Review related assets and product/project versions when they are relevant. A company’s earlier cases can explain a recurring issue that is not obvious from the latest contact’s message.Example: a month-end blocker“Exports are failing” leaves several questions open. Capture which subsidiaries are affected, when the failure began, the selected ledger, the error, and the customer’s deadline. Keep observations separate from your suspected cause. Then use the relevant case and customer history to avoid repeating work already completed. Make the next useful update Use a customer reply when the customer needs an answer, a question, or a progress update. Check recipients and the reply address before sending.Use an internal activity or discussion for internal coordination. Confirm the audience rather than assuming every editor is private.Make the case state reflect the work and confirm its owner or queue. A note that says “sent to engineering” is not a substitute for an actual handoff.Record what was tested, what remains unknown, and the next action. Include useful evidence without unnecessary private data.When the solution is reusable, search for existing knowledge, reuse or improve it, and follow the review/publication process for new content.Drafts need care: Case Add’s Save draft creates a case with a draft first activity, and lifecycle processing can still occur. Read the focused creation guide before using it as an unsent starting point. Finish the shift with a clear handoff Revisit your assigned work and the queues you cover. Identify commitments that extend beyond your shift.For each handoff, record the customer impact, checks completed, current hypothesis, next action, responsible person/team, and next update time.Use the permitted assignment or escalation action and verify the resulting owner or queue.Make the next customer update explicit. A specialist handoff should not leave the customer wondering who will respond.Record planned time away in the appropriate calendar and arrange coverage. An absence event does not automatically reassign every case. A supervisor’s onboarding check Ask a new colleague to find an assigned case, explain its customer and business impact, identify the next action, locate the relevant guide, and show how they would hand it over. Use a training example or read-only review before sending a real customer message. Confirm role access with the administrator if an action is unavailable.A useful daily review looks at actionable work and exceptions. Use scorecards for patterns over time, and case views for deciding what needs attention now. Avoid using a single percentage as a substitute for understanding complex customer work. Your first ten minutes in the agent workspace Sign in with your own agent account. If verification is required, complete the code step using your saved method. Confirm that you are in the intended Supportbench account.Open Cases. Select yourself in Agent where appropriate, then check Queue and View. An empty list can mean the selected scope has no matching work; it does not necessarily mean there are no cases.Open a case by its visible ID or subject. Use a training example or read-only review during onboarding. Locate the company/contact, current owner or queue, state and newest customer activity.Read backwards far enough to understand the latest promise and outstanding question. Ask your lead to explain your team’s status names, priority rules and service commitments; those settings can be configured.Locate the customer reply and internal activity controls without sending anything. Explain to your trainer which audience each uses, how you would check recipients, and where you would record an internal finding.Find the relevant knowledge article and check its prerequisites and customer access. Practice explaining the next action and who owns it.A service commitment is something support must deliver or follow up on. An SLA can define a measured response or resolution target under your configured policy. “I will update you by 15:00 UTC” is also an operational commitment even when it is not a separate SLA timer. Know where both are recorded and who is responsible. Leave a handoff another agent can act on Example internal shift handoff“Case c-12345, customer order export interrupted. Current owner: integration queue; covering agent confirmed. Verified: retry fails with the same validation error; logs and last successful timestamp are in the case. Open question: whether the required mapping field changed. Next action: specialist compares the two payloads. Customer update due 15:00 UTC; covering agent owns that update. Do not promise a release date until engineering confirms it.”Replace every example fact with the actual case evidence. The handoff is complete only when the receiving person or team knows about it and can access the case. A note alone does not assign ownership, create a task or send a customer update. Before leaving, verify the saved assignment and any due work using the focused case and task procedures. When your workspace looks different SituationNext stepA case is missingCheck agent/queue scope, the selected view, column filters, division access, and case permissions.A command is unavailableYour role or the record’s state may limit it. Ask the responsible administrator; do not use a different record to bypass access.Two agents see different countsCompare effective access and list scope before assuming the data is inconsistent.You are unsure whether content is publicCheck the activity type or article audience before saving or sending.A case keeps returning without progressReview ownership, the next action, customer dependencies, and whether escalation is needed. Continue learning Work a case from first response to resolutionBuild useful case viewsCoordinate with discussions and mentionsTurn support work into reusable knowledgePlan absence and coverage Was this article helpful? Yes No Thank you for your feedback.