Enterprise PPC Automation: The Draft-Approve Model Explained
Enterprise PPC teams are not looking to hand the keys to an AI. They are looking to eliminate the gap between when an account problem appears and when a human has time to act on it. Those are different problems, and they have different solutions.
The response-time gap
A campaign starts overpacing at 11 PM on a Thursday. The account manager does not log in until 9 AM Friday. By then, ten hours of budget have been consumed at an inefficient rate. No rule fires because the threshold was set for a 20% deviation and this one hit 18%.
This is not a failure of the account manager. It is a structural property of any system where a human has to observe before they can act. The response-time gap is proportional to the hours between sessions.
What the draft-approve model changes
SpendSignoff's always-on loop runs continuously inside a daily spend envelope. When it detects an overpacing pattern, it does not wait. It drafts a budget adjustment immediately — before 11:15 PM — and puts it in the approval queue.
The account manager arrives Friday morning to a queue of drafted changes, each with a before→after diff and a plain-English rationale. They approve, reject, or modify. The response-time gap is now the time it takes to review a queue, not the time until the next session.
The loop never spends unilaterally
mcp.draft scope, not mcp.approve. Drafts queue up. The live account does not change until a human approves. This is not a configuration option — it is a property of the scope model.Configuring the daily spend envelope
The spend envelope is the outer constraint on all loop activity. It defines the maximum daily budget change — in absolute dollars and percentage terms — that the loop is allowed to draft within a rolling 24-hour window.
Setting it too tight means the loop drafts changes too small to matter. Setting it too wide means the queue fills with large changes that slow human review. Most enterprise teams start at ±15% of current daily budget and adjust after observing the queue volume for two weeks.
- Absolute cap — e.g., no single draft can increase any campaign budget by more than $2,000/day.
- Percentage cap — e.g., no single draft can move more than 20% of the account's total daily budget.
- Category exclusions — brand campaigns and high-sensitivity ad groups can be marked as out-of-scope for loop drafts.
FAQ
- How is this different from automated bidding in Google Ads or Meta?
- Automated bidding adjusts bids at the auction level — within a campaign and strategy you configured. SpendSignoff drafts structural changes: pausing campaigns, reallocating budgets across campaigns, adjusting daily caps. Different layer, different kind of decision.
- What if the loop produces too many drafts to review?
- The sensitivity threshold is configurable. You can set a minimum impact size (e.g., only draft if the change affects more than $500/day) to reduce queue volume. You can also configure digest mode, which batches related drafts into a single review item.
Connect an account read-only and watch the operator work.
Reads are free on every plan. Nothing spends without your two-step approval.
Related reading