Ads Control
Ads Control
Checking…
Mode
AGENCY COMMAND CENTER

Cross-platform portfolio

Rank Google Ads and Meta Ads accounts by attention needed before opening an individual account.

Accounts scanned—Google + Meta
Critical—Review first
Needs attention—Material deterioration
Scaling—Efficiency looks expandable
Healthy / watch—No major flag
PORTFOLIO TRIAGE

Accounts needing attention

Click Scan portfolio to rank a limited set of accounts. Keep scan limits conservative during development.
PORTFOLIO AI

Ask across Google + Meta

Run a portfolio scan, then ask which accounts need attention, where CPL is deteriorating, or which accounts look scalable.
SAVED INVESTIGATIONS

Recent AI work

Recent account conversations will appear here.
Multi-channel foundationGoogle and Meta intelligence remain read-first. Campaign creation and routine Operations changes use separate human-approved write gates and can remain fully disabled in production until you enable them.
AGENCY OPERATIONS & ZERO-MISS QC

Operations Center

Run the same health, structural, fatigue and pacing checks across every ad account. Human decisions remain mandatory for changes.

Consistent checksSaved client QC policies control thresholds; missing API-only signals are shown as unavailable rather than guessed.
Accounts scanned—Google + Meta
Critical accounts—Review first
Needs attention—Warnings
Healthy—No configured breach
QC issues—Critical + warning
MORNING HEALTH CHECK

Priority issues

Run a check to begin.
Run Morning QC for daily delivery/performance/pacing checks, or Weekly QC for structural and fatigue monitoring.
Human-in-the-loop operationsOperations Center identifies and prepares routine changes. Budget or pause actions require explicit approval, exact confirmation and live-state revalidation. Live writes remain off unless platform-specific backend flags are deliberately enabled.
ACCOUNT COMMAND CENTER

Google Ads overview

Choose an account and run analysis to load the decision layer.

PERFORMANCE

Performance trend

Daily performance
Run analysis to load trend data.
AI OPERATING BRIEF

What needs attention

The daily action plan will summarize the most important account movement here.

PERFORMANCE SNAPSHOT

What is driving the account?

Run analysis to load performance data.
PERFORMANCE EXPLORER

Performance by entity

Move from account totals to the campaigns, products, keywords, queries, ads and PMax groups creating the result.

CAMPAIGN PERFORMANCE

Campaigns

Open this page after running account analysis.
BUSINESS & SITE TRUTH

What the business actually got

Ad-platform claims next to GA4 site behaviour and Shopify orders for the selected client. Each source stays separate; gaps are shown, not hidden.

ATTRIBUTION CHECK

Platform claims vs GA4 vs Shopify

Choose a client and run analysis, then open this page.
AGENTIC PERFORMANCE INTELLIGENCE

AI Copilot

Ask anything about the selected account. Copilot plans a bounded read-only investigation across platform reporting, Ads Control diagnostics and Official Meta Ads MCP when relevant.

Objective: run analysis Date range Evidence-backed · read onlyBounded tool planning
✦
Ask Copilot to investigate the account.

Try “Why did ROAS fall?”, “What changed recently?”, “Check creative fatigue”, or “Find wasted spend.”

AI Copilot · read only
AI CAMPAIGN BUILDER · HUMAN CONTROLLED

Build a campaign with AI

AI proposes the setup. You edit it. Ads Control validates it. Nothing is created until all three approval gates pass, and every created entity starts PAUSED.

No draft
Mandatory safety flowStrategy approval → Ads & tracking approval → Validation → Final approval → type CREATE PAUSED → create. No automatic activation.
1 · CAMPAIGN REQUEST

Tell AI what to build

2 · REVIEW & EDIT

Campaign draft

—
Generate a campaign draft. Every generated field will be editable before approval.
3 · MANUAL APPROVALS

Three human checkpoints

CREATE AS PAUSED ONLY
Gate 1Campaign strategyObjective, structure, budget, bidding and targeting.
Gate 2Ads & trackingCreatives/copy, keywords, URLs, pixel/event and UTM.
ValidationIndependent technical checksRequired fields, IDs, budgets, RSA limits and PAUSED-state enforcement.
Gate 3Final creation previewApproves the exact current draft fingerprint after validation.
Validation results will appear here.
4 · CREATE

Final controlled creation

Complete all approval gates first.
No activation call exists in Campaign Builder V1.
APPROVAL WORKFLOW

Action Center

Review optimization recommendations here even when no executable mutation exists. Build execution proposals only for changes that can be safely previewed and approved.

✓
Human approval gate is active.

Review-only recommendations stay visible here; execution-capable changes use before → after previews and human approval.

Run analysis to populate the review queue. Use Build execution proposals when you want budget/negative-keyword changes prepared for approval.
DECISION INTELLIGENCE

Optimization Plan

What the system found, why it matters, what to do next, and the expected impact.

Fix efficiency0
Scale safely0
Reduce waste0
Monitor / recover0
How to use this pageThe system automatically chooses the right KPI logic for the selected account.
PriorityTypeWhat we foundSuggested next stepExpected impactConfidence
Run analysis first.
PREDICTIVE INTELLIGENCE

Forecasts & scenarios

Forward-looking advertising-performance estimates with confidence ranges. These are scenarios, not guaranteed outcomes.

ModelBounded trend + run-rateDiminishing marginal efficiency on spend increases
Run a forecast to load projected performance.
SCENARIO SIMULATOR

Budget impact

Modeled · not guaranteed
Choose a budget scenario and run the forecast.
CAMPAIGN OUTLOOK

Campaign forecasts

No forecast loaded.
PRODUCT OUTLOOK

Product forecasts

No product forecast loaded.
GOVERNANCE

Change History & Audit Log

A record of dry runs, validations and executed changes for the selected client account.

Why keep this?It gives you accountability and troubleshooting context: which proposal was used, what mode ran, when it happened, and the execution ID. Later this is also the basis for rollback and post-change monitoring.
No change history loaded.
ADMIN

Team

Add people, set their role, and choose which clients and accounts they can use. New people start with no access.

NameRolePasswordAccess
CLIENT-SPECIFIC GUARDRAILS

Client Settings

Each ad account has its own protection rules and fiscal calendar. Switching accounts loads that platform/account's saved profile.

Not saved yet
GLOBAL SAFETY

Execution capability

These are backend-level limits and apply across every client.

SELECTED CLIENT

Client policy

These settings are saved separately for the account selected above and are used when generating action previews.

Objective detectionRun analysis to detect automatically
Operations Center QC policy
AI Analyst periodsUsed when a prompt says “financial year”, “FYTD”, or compares fiscal years.