The ecommerce data layer for AI

Connect Google Ads to Claude and ChatGPT

Google Ads knows your spend and its own attributed conversions. It does not know your cost of goods or your fulfilment bill. MerchantFlow supplies both so your AI can answer in profit, not just ROAS.

What you can ask once it is connected

Which Google campaigns had the best return last month?

Branded search returned 6.2x on A$310 spend. Shopping returned 3.8x on A$1,940. Performance Max returned 2.1x on A$820 and was the only campaign below your break-even ROAS of 2.4x for the period.

Example answer, shown to illustrate the shape and level of detail. Your figures come from your own connected account.

What Google Ads alone cannot tell you

Connecting an AI directly to Google Ads is genuinely useful, and for some questions it is all you need. It stops being enough the moment the question involves money you spent somewhere else.

  • Cost of goods and landed cost behind each converted order
  • Fulfilment, shipping and payment fees on those orders
  • Spend on your other platforms, which shapes the blended picture
  • Refunds against orders Google already counted as conversions

What MerchantFlow adds

  • Conversion value taken from Google's own reported figure rather than reverse-engineered from ROAS
  • Per-country spend resolved through several fallbacks, because Shopping and Performance Max campaigns frequently do not appear in Google's standard geographic report
  • Spend converted from micros into your store currency, at the rate for that day
  • Multiple customer accounts under one connection, each able to be paused independently
  • Every day re-anchored to your store's timezone, because Google reports in the ad account's
  • Campaign types split into Search, Shopping and Performance Max, with branded search separated out

MerchantFlow reports spend, attributed revenue and ROAS at campaign level. Profit is blended across the business rather than attributed per campaign. Campaign-type grouping is inferred from your campaign names, so a consistent naming convention makes it more accurate. Google Ads access also requires a developer token, which MerchantFlow supplies - you do not need your own.

How to connect it

  1. 01

    Connect Google Ads to MerchantFlow

    Authorise Google Ads from your MerchantFlow integrations page. We start pulling and normalising the data straight away.

  2. 02

    Add MerchantFlow as a connector in Claude or ChatGPT

    Approve read-only access from your AI client's connector settings. Nothing to install on your machine. Your dashboard shows the current steps for whichever client you use, because the clients move their menus often.

  3. 03

    Ask in plain English

    Your AI now reads the same figures as your dashboard, joined across every source you have connected.

Included on every paid plan, trials included. Starter reads 90 days of history, Pro reads 12 months, and Plus and Agency read your full history.

The tools your AI calls

These are the actual tools MerchantFlow exposes for Google Ads questions. Every claim on this page maps to one of them.

get_ad_performanceget_channel_roasget_pnl_summary
About the Google Ads integration

Common questions

Can it change anything in Google Ads?

No. The connection is read-only and no write access exists to grant. Your AI can read figures and reason about them; it cannot create, edit, pause or delete anything.

Will the numbers match my dashboard?

Yes. Your AI calls the same calculations the MerchantFlow dashboard uses, so the same question over the same period returns the same figure in both places.

Connect Google Ads and ask your first question

Start a free trial, connect your sources once, then ask Claude or ChatGPT what actually happened.

Start free trial