NotFairNotFair
Start now

NotFair · Google Ads guides

A Google Ads AI tool built for people who still approve the work

NotFair connects live Google Ads data to your AI client so analysis, recommendations, and approved edits happen in one workflow instead of across exports, dashboards, and spreadsheets.

Updated September 5, 2026 · NotFair Team

  • Diagnose campaigns, keywords, search terms, budgets, and recent changes from chat
  • Use Claude or Codex to draft concrete edits instead of generic recommendations
  • Approve writes before they reach Google Ads and keep a change trail afterward
Connect Google Ads

Choose the tool for the work you need

Google’s native tools handle auction-time bidding. PPC management tools help with cross-account rules and oversight. Creative tools generate assets. NotFair connects live account data to a conversational assistant so you can investigate performance and review supported changes. These tools serve different jobs and can work together.

The first workflows to run

Start with high-spend zero-conversion campaigns, search-term cleanup, weak Quality Score pockets, budget reallocation, and recent-change review. Start with one clearly defined problem, check the evidence, and review a narrow change before applying it.

What the tool needs to know before it can be useful

A Google Ads AI tool is only as good as the context it can inspect. NotFair gives the agent structured access to account hierarchy, campaigns, ad groups, keywords, search terms, spend, conversions, recommendations, and recent changes. That lets the answer move past generic advice like improve ad relevance and into account-specific reasoning such as which campaign spent without conversions, which match types expanded too far, or which budget is blocking a working segment.

  • Account structure and accessible customer IDs
  • Campaign, ad group, keyword, search term, and asset performance
  • Recent changes that explain sudden CPA, CPC, or volume movement
  • Reviewable write paths for negatives, bids, budgets, ads, and campaign state

How NotFair differs from reporting dashboards

Dashboards are useful when you already know which metric to inspect. NotFair is built for the messy middle: you know performance changed, but you do not know whether the cause is search terms, bidding, budget, conversion lag, network mix, landing pages, or recent edits. The agent can pull several slices of the account in one investigation and return a prioritized action list instead of another static chart.

  • Ask a plain-English diagnostic question and let the agent choose the account cuts
  • Turn the diagnosis into a concrete draft change instead of a detached insight
  • Keep final approval human-owned so the workflow stays useful for operators

The safest rollout path

The best first session is read-only. Ask the agent to explain account structure, summarize last-30-day performance, identify the largest waste pockets, and name the exact write it would do first. After that, approve one narrow change and review the before/after context. This prevents AI from becoming vague advisory software or unsafe automation.

  • First read: account map, spend, conversions, CPA, and change history
  • First recommendation: one high-confidence action with the evidence attached
  • First write: a narrow negative keyword, bid adjustment, or campaign-state update

Try a focused first task

CPA spike diagnosis

Why did cost per lead increase over the last 7 days? Check search terms, device mix, network mix, campaign budgets, and recent changes before recommending fixes.

A ranked explanation with the evidence for each cause and a small set of reviewed edits.

Zero-conversion spend audit

Find campaigns, ad groups, keywords, and search terms that spent more than $100 in the last 30 days with no conversions. Group them by likely action.

A cleanup queue for negatives, bid reductions, pauses, and landing-page review.

Launch structure draft

Create a search campaign structure for this offer using the current account naming style. Include ad groups, keyword themes, negatives to start with, and ad copy angles.

A campaign draft the operator can edit before any write reaches Google Ads.

Questions before you connect

The MCP server is the connection layer. The AI tool workflow is what that connection enables: live reads, recommendations, approved writes, and change review from an AI client.

No. NotFair is designed for agent-assisted execution. Reads can happen freely, but account changes are reviewable and approval-gated.

NotFair works with MCP-compatible clients including Claude, Claude Code, OpenAI Codex CLI, Cursor, Cline, and custom clients.

Next steps