Rehearsing a Delivery Interruption Before It Happens: A Tabletop Drill for Ad Teams

Editorial team Spend Ledger

Set up a realistic, low-stakes exercise

A delivery interruption is easier to manage when the team has already practised how to respond. A tabletop drill lets people talk through a hypothetical pause without changing live campaigns, contacting a platform or making account changes. The goal is not to predict every cause. It is to find out whether the team can recognise a problem, coordinate a response and keep an accurate record.

Keep the exercise focused. Choose a plausible scenario, such as a campaign that stops delivering or an account that becomes unavailable for advertising. Do not assume the cause. The interruption might relate to billing, a review, an account setting, a technical issue or a platform decision. The team should treat the scenario as unconfirmed until it has checked the relevant account information and official notices.

Invite the people who would actually take part in a response. Depending on how the work is organised, that may include campaign operators, an account owner, a finance contact, a client or business contact, and a person responsible for platform support. If an agency or another service provider manages some parts of the account, include a representative or clarify how that provider would be reached.

Make clear that this is a discussion, not a test of individual performance. The useful question is not “Who made a mistake?” but “What information would help us act correctly, and where would we find it?” This keeps the exercise practical and makes it easier to surface unclear responsibilities.

Choose a scenario and state what is known

Begin with a short scenario statement. For example: “A team member sees that expected ad delivery has stopped. The reason is not yet known, and the account’s current status needs to be checked.” Avoid adding dramatic details that steer the group toward a particular explanation.

Give participants only the information they would reasonably have at that moment. They can then describe what they would check first, who they would notify and what they would avoid doing before the cause is clearer. Introduce new information gradually, as if the team were receiving it during a real incident. This may reveal whether people know where to look and how they would handle uncertainty.

The exercise should not encourage workarounds. Participants should not propose opening replacement accounts to bypass a restriction, concealing connections between accounts, changing identity details or avoiding verification. If the scenario involves a policy decision, review or account restriction, the appropriate path is to understand the notice, address the stated issue and use the platform’s official appeal or support process when warranted.

Keep the hypothetical separate from live account data. Do not ask people to share passwords, sensitive payment information or private customer details in the discussion. The drill needs to test the response process, not expose confidential information.

Decide who notices and who owns the response

Ask how the interruption would first come to the team’s attention. It might be noticed during routine campaign monitoring, through an account notification or when a planned delivery check shows a change. The important point is to identify the normal detection route and any gaps in it. If nobody is clearly responsible for reviewing account status, that is a finding to record.

Next, establish who coordinates the response. One person should be responsible for bringing the relevant facts together, while others handle specific checks or communications. This does not mean only one person can investigate. It means the team should know who is keeping the response organised and preventing conflicting changes or duplicate messages.

Clarify who has authority to make account or campaign changes, and who can speak for the business in a support request. If an account is managed within a larger business structure, distinguish the responsibilities of the business-level owner from those of people working in individual ad accounts. A person who can edit campaigns may not be the person who can resolve billing, ownership or access questions.

Where an agency or provider manages the account, discuss the agreed division of work rather than relying on assumptions. The written arrangement should make clear who monitors delivery, who can access relevant notices, who contacts support, who communicates with the client and who retains records. If those responsibilities are missing or ambiguous, treat that as a documentation issue to resolve after the drill.

Walk through the first checks

Have the group describe the first checks in a sensible order. Start with the information available inside the relevant advertising account: current status, delivery indicators, notifications, campaign settings and any visible review or policy messages. Check whether the interruption affects one campaign, an ad account or a broader business-level structure, without assuming that one visible symptom explains another.

Then consider operational causes that the team can verify, such as whether a campaign is scheduled to run, whether its settings were changed, whether billing information needs attention or whether access to the account has changed. The group should distinguish confirmed facts from possibilities. “The account displays a notice” is a fact; “the notice must mean the account was restricted” may be an unsupported interpretation.

Agree how to capture relevant information while protecting sensitive data. A record might note the account or campaign involved, the time the issue was observed, the visible status, the checks completed and the person responsible for each next step. Do not copy sensitive payment details or unnecessary personal information into general incident notes.

Also discuss what should not happen during initial checks. Avoid repeated, uncoordinated edits made in the hope of restoring delivery. Do not change identity information to get past a check, or make changes that obscure the original state. If the team cannot tell whether an action is safe or permitted, pause and consult the platform’s official guidance or support channel.

Practise contacting official support

If the first checks do not explain the interruption, the team should know how it will use the platform’s official support channels. Practise identifying the appropriate route from within the platform or its official help and policy pages. Do not rely on unofficial contacts, guessed addresses or links copied from an unverified message.

Before making a request, gather a concise account of the issue. Include what the team observed, which relevant checks it completed, what notices appeared and what question remains unanswered. Keep the description factual and consistent. If the platform requests information or documentation, provide it through its official process and ensure it is accurate and authorised for sharing.

The exercise can test who is permitted to submit a request and who will track any response. If a provider manages the account, establish whether the provider or the business owner should contact support, and how each will share updates. Avoid sending multiple conflicting requests simply because the team is unsure who owns the task.

When a notice provides instructions or identifies a possible cause, the next step is to understand and address that issue through the stated process. An appeal is appropriate when the team believes a decision should be reviewed, but it should be accurate and grounded in the available facts. The drill should not script a particular outcome: support decisions and review processes belong to the platform.

Prepare communication for owners and clients

A delivery interruption may affect campaign plans, reporting or business expectations. The team should decide who needs to be told, when an update is appropriate and who is authorised to provide it. The message should separate what is known from what is still being checked.

A useful update can state that delivery has changed, describe the checks underway and explain the next planned communication through the agreed channel. Avoid presenting an unconfirmed cause as fact, suggesting that delivery will resume by a particular time or promising a specific result. If the status is uncertain, say so plainly.

For client work, agree in advance how the agency or account manager will explain the situation. The account owner may need to be involved if the interruption concerns account access, billing or a platform decision. If the team uses an agency-provided account, the arrangement should also cover who can share status information and who is responsible for keeping the client informed.

Discuss whether business decisions need to be considered while the issue remains unresolved. For example, owners may need to review spending plans, pause related work or adjust expectations. These are business decisions, not methods for bypassing a platform restriction. Any change should remain within the platform’s rules and the authority agreed for the account.

Record decisions, uncertainties and handoffs

During the exercise, someone should keep notes on the process rather than trying to capture every sentence. Record who was assigned each task, where the relevant information is found, which decisions require approval and what the team would communicate. Mark unresolved questions clearly instead of filling gaps with assumptions.

Include the handoffs. A response can fail even when each person understands their own role if nobody knows when the work passes from the campaign operator to the account owner, finance contact, provider or client lead. Describe what information should travel with each handoff and where the record should be stored so the next person can find it.

Review whether the team can access the materials it expects to need, such as account ownership records, internal contact routes, billing procedures and the current support process. Do not place credentials in the runbook. Access should be managed through the organisation’s approved security practices, and the runbook should point to the appropriate internal source rather than reproduce sensitive information.

At the end, ask each participant what was unclear or difficult. Pay attention to conflicting answers about ownership, missing access, uncertainty about the correct support route and differences between written agreements and actual practice. These observations are more useful than a simple conclusion that the team is prepared or unprepared.

Turn findings into runbook updates

Convert the discussion into specific changes to the incident runbook. If nobody knew who should coordinate the response, assign that responsibility and document the handoff. If staff did not know where to find official support guidance, add an instruction for locating the platform’s own current help and policy pages. If client communications were inconsistent, establish who drafts and approves updates.

Write procedures in plain language. A useful entry explains what triggers the response, what to check, who owns each task, what information to record and when to escalate. It should distinguish routine delivery checks from situations involving a notice, review or restriction. It should also remind staff to follow official instructions and avoid attempts to evade account rules, limits or verification.

Keep changeable platform details out of fixed internal instructions when they may become outdated. Instead, tell staff where to confirm current requirements on the platform’s official pages. The platform publishes current values and procedures on its own help and policy pages; the runbook should not guess at them.

Finally, make updates easy to maintain. Assign an internal owner to review the runbook when account responsibilities, provider arrangements or platform processes change. Note what was revised and why, without implying that the drill can prevent every interruption. A tabletop exercise is useful because it exposes practical gaps while the team can still improve its process, not because it can guarantee how a platform will respond.