Using the Ads Agent
The per-campaign Ads Agent and the cross-campaign Portfolio Agent — model and effort controls, playbooks, the research brief, and why every change still needs you.
Two chat assistants live inside GrowthAd Manager. The Ads Agent is scoped to one campaign and knows its metrics, ad groups and history. The Portfolio Agent reasons across the campaigns you select. Both analyse and propose. Neither executes anything against your spend.
Before you start
Both agents are owner-only and need the provider connected under SettingsIntegrations. The Portfolio Agent additionally needs at least one Google campaign selected — otherwise it says Select at least one Google campaign.
Which agent to open
| Ads Agent | Portfolio Agent | |
|---|---|---|
| Scope | One campaign | The campaigns you select |
| Where | The floating panel on a campaign detail page | Ad Manager → Portfolio Agent |
| Good at | Search terms, negatives, RSA copy, one campaign's CPA | Budget reallocation, comparing campaigns, pause candidates |
The controls
| Setting | What it changes | Default |
|---|---|---|
| Model | Which model answers. Models marked (analysis only) cannot propose campaign changes at all. | Default |
| Effort | How much reasoning to spend: Low, Medium, High, X-High or Max. Higher takes longer. | — |
| Web research | Whether the agent may look things up. Auto, On or Off. Forced to Off for analysis-only models. | Auto |
| Playbook | Inserts a saved long-form prompt so you do not have to write the brief yourself. | — |
| Campaign scope | Portfolio Agent only. Which campaigns the agent may read and reason about. | — |
Analysis-only models cannot propose changes
Pick one of those and you will get good reading of the data and no actionable proposals. The panel says so — Analysis only — this model cannot propose campaign changes. If you wanted a recommendation you can approve, switch model first.
Playbooks and quick prompts
Six playbooks cover the recurring jobs: Full campaign buildout, Search terms & negatives, RSA and landing pages, Competitor gap, Bidding and budget, and Change audit & learning.
Quick prompts are one click each: Analyze performance, Find wasted spend, Ad group breakdown, Keyword analysis, Demographic insights, Campaign settings, Google recommendations, What changed recently? and Budget simulation. Condition-aware ones appear when they apply — Improve low CTR, Fix zero conversions, Reduce high CPA.
The Portfolio Agent has its own set: Portfolio overview, Budget reallocation, Find wasted spend, Compare top campaigns, Pause recommendations and Growth opportunities.
The campaign research brief
This is the highest-value thing on the page and the most frequently skipped. It is structured business context the agent reads before it answers anything, and the difference between a generic answer and a useful one.
| Field group | What to put in it |
|---|---|
| Services | What you actually sell — the placeholder suggests mobile detailing, tree service, vacation rentals |
| Service areas | Where you serve |
| Competitors | Who you lose to |
| Differentiators | Why someone picks you |
| Services not offered | So the agent stops bidding on them |
| Poor-fit traffic to block | The searches that waste money |
| Compliance-sensitive wording | What you may not say |
| Preferred and avoided landing pages | Where traffic should and should not go |
| Positioning, ICP and audience detail | Who you want |
| Proof points | What you can claim |
| Price posture | Premium, mid, value |
| Brand voice | How ad copy should sound |
| Seasonality | When demand moves |
Services not offered earns its keep immediately
Most wasted spend in a small account is traffic for something the business does not do. Filling that one field turns find wasted spend from a guess into a list.
Approval, always
The agent proposes. You approve. There is no path where a chat message moves money on its own.
No autonomous optimisation runs by default
Durable Google Ads optimisation tasks — the research, plan, approve, execute, verify loop — ship behind a flag that is off by default. With it off, chats behave like ordinary approvals and create no runnable tasks. On the Meta side, the auto-approve rules card stores rules that nothing evaluates: every action still requires a manual approval. Do not plan around unattended changes.
Approved changes go through the same staged review and the same frozen money guards as an edit you made by hand.
History
The panel keeps conversation history and a separate action history, so you can see what was proposed, what you approved, and when. Use the action history before asking what changed recently? — it is the record of your own decisions, not the provider's change log.
Troubleshooting
| Symptom | Cause | Fix |
|---|---|---|
| The agent explains but never proposes anything actionable | An analysis-only model is selected | Switch model; web research is also forced off for those |
| The Portfolio Agent refuses to start | No campaigns selected | Select at least one Google campaign in the scope picker |
| Answers are generic and miss your business | The campaign research brief is empty | Fill it in, especially services not offered and poor-fit traffic |
| An approved change did not reach the provider | Approval and application are separate steps | Check the staged-changes bar and press Apply |
| A proposal was rejected as too large | It crossed a frozen money guard | Step it up over smaller edits |
| A Meta approval stopped working overnight | Meta approvals expire after 24 hours | Ask the agent to re-propose |
| You expected a rule to act while you were away | The Meta auto-approve rules card is dormant | Every action needs a manual approval |