Google Ads for AI agents.
Stop paying for clicks that never convert. Your agent reads the account every day, blocks the search terms that only spend, and waits for your approval on anything your policy doesn't cover.
✓The audit is free and read-only. Nothing in your account changes.
- gads_get_sync_status · synced 2h ago · 90 days of data
- gads_get_optimization_plan · CPA $64 against a $55 target
- gads_get_search_terms · 41 terms spent $1,890 with no conversion
- gads_propose_add_negative_keywords · 41 negatives · applied under your policy
- gads_propose_budget_change · +40% on Brand · waiting for your approval
87
MCP tools
Every dashboard action, callable by your agent.
30%
Largest budget step without you
Anything bigger waits in your approval queue.
20
Auto-applied changes a day
The ceiling per account, in any rolling 24 hours.
0
AI credits per agent call
Your agent runs on your own Claude or ChatGPT plan.
[ 01 / 06 ]Agent + approval queue
Your agent proposes. Your policy decides.
Every change your agent makes is a proposal first, with what moves, from what to what, and the numbers that triggered it.
- Bounded fixes, like a negative keyword or a budget trim, apply under the policy you set.
- Anything larger waits in the approval queue with the rule it tripped named on the card.
- Every proposal is re-checked when it applies. If the account drifted, it goes back to the queue.
- Applied changes revert in one click, from the state saved before anything moved.
[ 02 / 06 ]For your agent
Connect your agent in one command.
No API key to paste. Claude Code, Cursor, Codex and claude.ai use the same URL and sign in through the browser.
Then ask — "Read my account's optimization plan and tell me where the spend is leaking."
claude mcp add --transport http bidandscale https://bidandscale.com/api/mcp[ 03 / 06 ]Smart Bidding
You don't fight Smart Bidding. You feed it.
Google optimizes inside the budget you set, toward the signal you gave it. Your agent works on the inputs the auction can't fix for itself.
- Search terms nobody would have bought, still spending every day.
- Match types too loose for the budget behind them.
- Budgets set by habit instead of pacing against a target.
- Conversion tracking that double-counts, so the target is wrong.
Key takeaway
The auction is Google's job. The inputs are your agent's.
[ 04 / 06 ]Why Bid & Scale
Built for agents, not bolted on.
Getting data
GAQL queries against a live, rate-limited API
A synced warehouse, plus GAQL when you need it
Making a change
A mutate goes live the moment the agent sends it
A proposal, applied by your policy or your approval
Undoing a change
Rebuild the old settings by hand
One-click revert from the state saved before it moved
Knowing what happened
Google's change history, with no reasons
A ledger of every change, its reason, and who approved it
Strategy
Whatever the prompt says this time
Your targets, pacing and CPA band, read on every plan
[ 05 / 06 ]What your agent does
The whole account, not just reports.
It can read everything. Every change it makes is a proposal, logged with its reason.
Block wasted spend
Mine search terms, harvest negatives, and tighten match types.
Work the optimization plan
Pacing against your targets, and a ranked list of what to scale or fix.
Move budgets and bids
Budget changes, target CPA and ROAS steps, and keyword bids, inside your caps.
Write and test ads
Responsive search ads and assets, staged paused when you want a review.
Run experiments
Set up, read and end campaign experiments, including AI Max tests.
Explain what changed
Google's change history and our own ledger, with the reason behind each change.
[ 06 / 06 ]Pricing
Free to read. You pay when changes apply.
The audit, the plan, and every proposal are free. A subscription lets approved changes apply, and its bill follows the spend we read from your account.
Free for agents
Read everything. Propose anything.
$0
No card. However many accounts you connect.
- A full audit: six areas scored on 90 days of data
- Every read over MCP and the API, from any agent
- Proposals drafted and queued — nothing applies
- A share link and a PDF you can send as-is
- Free for as long as you stay read-only
Copilot
Per ad account. Approved changes apply.
$299 /mo
$49 + 1% of $25,000
1.2% of your spend, all in
- Bounded fixes apply nightly, with the reason attached
- Everything larger queues for your approval
- Search-term mining and negative harvesting
- One-click revert on every applied change
- Full MCP access from Claude, Claude Code, or Codex
- Never bills over $499/mo
Spend is read from the account you connect, so your bill tracks your real Google invoice — nothing self-reported. Reads, the audit, the plan and every proposal are free, for you or your agent; Copilot is billed per ad account and is what lets approved changes apply.
Questions people ask us
Stop paying for attendance. Start paying for changes.
Connect your agent, run the free audit, and approve the first fix.