Try getting Google Search Console data into an AI chat the “proper” way and see how far you get. Create a Google Cloud project, enable the API, configure OAuth consent, generate credentials, work out which scopes you actually need. I can do that. Most marketers I know can’t, and honestly, why should they have to. connector.wtf skips that whole detour, and that’s the entire pitch.
In short
- The average marketer trying to connect an ad or analytics account to AI hits a Google Cloud project before they hit anything useful
- Everything on connector.wtf is read-only: the OAuth scope physically can't touch a bid or pause a campaign
- Most of the essential stack is live already, Google Ads, Meta Ads, LinkedIn Ads, LinkedIn Ad Library, HubSpot, Search Console, Bing Webmaster, WooCommerce, with GA4 on the way
Why this is harder than it should be
Every MCP server needs a way to authenticate against the platform it’s reading from, and for most ad and analytics platforms that means a developer console built for engineers. I’m fine in that environment, ten years of in-house marketing means plenty of API keys and service accounts along the way. But I’m also aware that’s exactly the kind of task that quietly kills a good idea, because a marketer looks at “create a Google Cloud project” and closes the tab. That gap, not the AI part, is the actual barrier to AI-enabled marketing ops for most teams.
What connector.wtf actually does
Sign in with Google. Pick which accounts the AI is allowed to see, manager accounts unlock their sub-accounts automatically. Paste the resulting MCP URL into Claude or ChatGPT as a connector. That’s it, no separate token to guard, no console to configure.
The read-only part is the bit I’d actually stake a client relationship on. The scope can’t change a bid, pause a campaign, or edit a single thing, it’s baked into what the token is allowed to request in the first place. For a marketer touching multiple ad accounts, that’s not a nice-to-have, it’s the reason I’d hand this to a client’s in-house team without a second thought.
What’s live right now
Google Ads is the deepest connector: account summaries, wasted spend reports, period comparisons, budget pacing, geo and device breakdowns, even raw GAQL for anyone who wants to write their own queries. Meta Ads pairs the official Meta Ads MCP into the same chat, so one agent ends up speaking Google and Facebook at once. LinkedIn Ads mirrors the Google Ads connector with B2B-specific pivots by seniority, industry and company size, and there’s a separate LinkedIn Ad Library connector for pulling competitor ads without needing an account at all.
Beyond paid media: Search Console, Bing Webmaster, HubSpot and WooCommerce are all live and all read-only. GA4 is marked “coming soon,” which is the one gap I’d actually want closed next, since most of my reporting still pairs ad spend against Analytics.
I’d point any marketer stuck at “I know AI could help me with reporting, but I have no idea how to connect my ad accounts to it” straight at connector.wtf. It’s the missing middle step between wanting an AI-enabled workflow and actually having one. It’s like having a whole team at your disposal.