AI assistants and search decision
NotFair
NotFair is open-source (official GitHub) so a technical user can self-host the repo to obtain the core MCP connectors and workflow instead of paying; running it requires multi-week setup and ongoing maintenance for integrations and auth.
Visit website↗$79/mo
$948/yr
Read off the official pricing page.
$20one-off6 h to build
$200/mo6 h/mo upkeep
On cash alone, building overtakes the subscription at 3 seats.
Open-source builds that already do this
Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need - the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship. All NotFair alternatives, with the arithmetic →
What a replacement has to do
- Agent reads live ad & analytics data, runs diagnostics to find wasted spend or tracking gaps, drafts prioritized fixes, shows a diff for each suggested write, and applies approved changes via Ads APIs with undo support.
What it still won’t have
- Hosted MCP servers and managed uptime/SLA
- Managed connector onboarding and hosted connector URLs (e.g. one-click Claude connector)
- Priority support and managed account/ads-strategist services
- Managed scaling, backups, and security patching provided by the vendor
What remains hard
- Brand trust
Part of the Claude Partner Network
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 3 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a self-hostable NotFair MCP server using Node.js + TypeScript, Express for MCP HTTP endpoints, Postgres for storage, Redis for job queueing, and Docker for deploy. Implement: 1) OAuth2 flows and secure token storage for Google Ads, Meta Ads, X/LinkedIn, Search Console, and GA4; 2) MCP-compatible endpoints that return reads (metrics, search terms, quality scores) and accept approve/diff write requests; 3) diagnostics engine that calculates spend, conversion, and spend-at-risk rules and returns prioritized fixes; 4) diff-and-apply write logic that prepares provider API changes, applies them after approval, and records reversible change history; 5) a minimal web UI or webhook endpoint to display diffs and accept approvals; 6) background jobs for weekly reviews and retry logic; Out of scope: managed hosting, proprietary Claude integrations, and paid ads-strategist services. Include thorough error handling, unit and integration tests, and Docker Compose and production deployment docs.
How we checked
How the score was reached
- Self-host verdict base92
- An open-source build was found+5
- 5 cited sources+3
- Price verified on pricing page+3
- Evidence score99
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.
How scoring works →Cited sources · 5
Every page the run actually retrieved.
- official productNotFair MCP — AI Ads Agent for Google Ads, Meta Ads & More
- official pricingNotFair Pricing | Free MCP Connectors & $99 Growth Plan
- official docsGoogle Ads + AI use cases | NotFair
- open sourcearc53/DocsGPT
- open sourcedzhng/deep-research
Integrity checks
What held up, and what did not.
