A new support agent can learn your tools quickly and still struggle to make the right decision. The missing piece is often the knowledge your existing team carries in its head: which promise was made, when a refund needs approval and who can unblock a delayed order. A useful handover brings those decisions into the open. Start with a small, usable playbook and test it against real situations.
Map the journey and the common questions
Walk through the journey from purchase or registration to delivery, use and follow-up. At each stage, write down what a customer might ask, which records the agent should check and what an acceptable next step looks like. Start with the questions your current team sees most often. For a store, that might include delivery tracking, address changes and returns. For a software business, it might be account access, plan changes and product navigation. Link each situation to the relevant source of truth. A short map with working links is easier to use during a conversation than a folder of disconnected screenshots.
Turn policies into decisions
A policy tells the agent what your business allows; an example shows how to apply it. Pair refund, replacement, cancellation and access policies with straightforward cases and a few exceptions. State the agent’s authority clearly: what can they approve, what must be checked and who decides when the standard policy does not fit? If a rule changes by market or product, put that distinction next to the rule. Avoid asking agents to infer it from an old ticket. Give every document an owner and a review date so the team can identify the current policy when two sources disagree.
Give every escalation an owner
“Ask operations” leaves too much unanswered. Name the responsible person or team, the approved channel and the information needed to make a decision. A useful escalation contains the customer’s issue, relevant identifiers, checks completed, actions already taken and the exact decision requested. Agree on what the agent should tell the customer while waiting, including when to follow up if there is no answer. Set an internal route for urgent situations and another for routine exceptions. Ownership should remain visible after the handover; sending a message to another department does not, by itself, close the customer’s case.
Practice before giving live authority
Choose anonymized examples that cover a routine inquiry, an ambiguous policy and a difficult conversation. Ask the agent to explain their checks and reasoning, not just draft a reply. Use a sandbox where possible, or a supervised session with carefully limited permissions. Confirm that the agent can find the right records, recognize an exception and leave useful notes. Test access separately from knowledge: being able to log in does not prove someone can safely change an account or issue a refund. A readiness review should identify remaining training needs and specify which actions still require approval.
Use early questions to improve the playbook
The first weeks will reveal gaps you could not see from the handover alone. Keep a shared list of repeated questions, unclear policies and cases that took unnecessary back-and-forth. Review it with the people who can approve changes. Sometimes the answer is coaching; sometimes the workflow or documentation needs fixing. Update one agreed source and tell the team what changed, rather than circulating another version by email. Make documentation maintenance part of the regular service review. A good playbook becomes more useful as the team works, and helps the next person get up to speed with less guesswork.