LinkedInSet the caps before launch
- Agree one monthly ceiling per client in writing, and turn it into the daily figure it implies.
- Set a limit on each campaign that the account cannot pass without a second approval.
- Keep test budgets apart from scaling budgets, so one test cannot drain the main line.
Set the guardrails
- Alert at 80 percent and again at 100 percent of the monthly ceiling, sent to a named person, not to a shared inbox that nobody reads.
- Watch pacing: compare spend to date with the share of the month that has passed. The budget pacing guide shows how to read it.
- Write the spike rule in plain numbers, for example spend above 1.5 times the trailing daily average, and check it every day.
Made-up scenario: a client approves 9,000 for March, which is 300 a day. On day 11 spend is 4,100, about 46 percent of the budget at about 35 percent of the month. The pause happens that afternoon, not at the month end.
Decide who can pause
- Name one person per client who may pause spend without asking, and one backup.
- Write the pause rule: what triggers it, such as ROAS under break-even for a set number of days, which campaign pauses first, and what must be true before a restart.
- Restart only with the same approval the launch needed. A restart with no review is a launch with no check.
- Log every pause and restart in the change log, with the reason and the time.
Review the caps each month
- Compare the caps with actual spend and with the approved plan. Change a cap only with a written reason.
- Check that no alert or guardrail was switched off during a busy week and left off.
