SOP template and free SOP generator

# [Process name] SOP number: [ID] Version: [1.0] Owner: [role] Last reviewed: [date] ## Purpose [One sentence: why this procedure exists and what outcome it protects.] ## Scope Applies to: [teams, products, regions]. Does not cover: [exceptions]. ## Roles and responsibilities - [Role]: [what this role does in the procedure] - [Role]: [approves, reviews or is informed] ## Before you start - Access to: [tools or permissions] - Inputs: [documents, data, requests] ## Procedure 1. [Action, written as an instruction.] 2. [Action.] a. [Sub-step, if the action has parts.] 3. [Action.] ## Escalation If [condition], contact [role] through [channel] within [time]. ## Records Save [what] in [where] for [how long]. ## Review [Owner] reviews this SOP every [period] and after any incident that involves it.

# [Task name] checklist Owner: [role] When to use: [trigger] - [ ] [Check or action] - [ ] [Check or action] - [ ] [Check or action] - [ ] [Check or action] - [ ] Record the result in [where] If any item fails: [what to do and who to tell].

# Ticket triage ## Purpose Every new ticket gets a priority, a category and an owner within [time]. ## Procedure 1. Open the oldest unassigned ticket in [helpdesk]. 2. Read the whole message, including attachments. 3. Set the category: [billing / technical / account / feature request / other]. 4. Set the priority: - Urgent: the customer cannot use the product, or there is a security or data risk. - High: a main feature fails and there is no workaround. - Normal: everything else. 5. If an article in the help center answers it, reply with the link and a one-line summary. 6. Otherwise assign it to [queue or person] for that category. 7. For Urgent tickets, also post the ticket link in [channel]. ## Escalation An Urgent ticket without a reply after [time]: tell [role].

# Refund request ## Scope Refund requests by email, chat or form. Chargebacks follow [other SOP]. ## Procedure 1. Find the order or subscription in [billing tool] with the customer's email. 2. Check the refund policy: [policy link]. Note the purchase date and the plan. 3. If the request is inside the policy, issue the refund in [billing tool]. 4. If it is outside the policy, ask [role] for approval before you reply. 5. Reply to the customer with the amount, the payment method it goes back to, and when to expect it: [typical days]. 6. Tag the ticket "refund" and add the reason in [field]. ## Records The billing tool keeps the refund. The ticket keeps the conversation.

# Escalating a bug to engineering ## Before you start Try to reproduce the problem with a test account. ## Procedure 1. Search [issue tracker] for an existing issue. If one exists, add the ticket link and stop. 2. Create an issue with: - Steps to reproduce, numbered - What happened and what should happen - Account or workspace ID (never passwords or full card numbers) - Browser, app version, screenshots - How many customers are affected 3. Set the severity with the same rules as ticket priority. 4. Link the issue in the ticket and tell the customer the problem is with the team, without promising a date. 5. When the issue closes, reply to every linked ticket.

# Customer complaint ## Procedure 1. Reply within [time] to say you have read it and who will handle it. 2. Restate the problem in one sentence so the customer can correct you. 3. Find the facts: account history, earlier tickets, order details. 4. Decide the remedy with the options in [policy]. Above [amount], ask [role]. 5. Reply with what happened, what you will do, and by when. 6. Log the complaint in [where] with the category and the remedy. ## Review [Role] reviews the complaint log every [period] for repeat causes.

# Account access and password reset ## Procedure 1. Never change an email address or reset a password from a ticket alone. 2. Ask the customer to use the reset link on the sign-in page first. 3. If they no longer have access to the email: verify identity with [method], such as the last invoice number. 4. If verification fails, do not change the account. Tell the customer what they can do next. 5. If it passes, make the change in [admin tool] and note what you checked in the ticket. ## Escalation Several failed verifications for one account: tell [security contact].

# Chatbot to human handoff ## Purpose A visitor the bot cannot help reaches a person without repeating themselves. ## Procedure 1. Open the handoff in [inbox] and read the full conversation first. 2. Greet the visitor by name if known, and say you have read the chat. 3. Answer the question, or ask one specific follow-up question. 4. If the answer should have been in the help center, note the question in [list] for the docs owner. 5. Close the conversation with the resolution category. ## Review Every [period], the docs owner turns the most common handoff questions into help articles.

# Help center article update ## Trigger A product change, three or more tickets on the same question, or a wrong answer found in an article. ## Procedure 1. Find the article, or create one if none covers the question. 2. Put the answer in the first two sentences. 3. Use the exact names of buttons and pages as they appear in the product. 4. Add one screenshot only where words are not enough. 5. Ask [role] to check the facts. 6. Publish, and link the article in the tickets that triggered it. 7. Set a review date in [where].

# Outage communication ## Procedure 1. Confirm the outage with [on-call role]. Do not post before it is confirmed. 2. Post on [status page] within [time]: what is affected, since when, and that the team is working on it. 3. Add a saved reply in [helpdesk] that links the status page, and tag related tickets "outage". 4. Update the status page every [interval] until it is fixed, even with no news. 5. When it is fixed, post the resolution and reply to all tagged tickets. 6. Within [days], publish a short summary of the cause and what will change.