How to Connect Google Ads to ChatGPT (1-Min Setup via Windsor.ai Native App)

You can connect Google Ads to ChatGPT in two no-code steps: connect your Google Ads account in Windsor.ai, then add the Windsor.ai plugin in ChatGPT. ChatGPT then reads your campaign, keyword and search term data when you ask. Google’s own route is an open-source MCP server that you run yourself; it has no tools to pause campaigns or change budgets and bids.

The account is full of data: spend, impressions, conversions, Quality Scores, search terms, device splits, and hundreds of other metrics. Yet, getting a clear picture of what’s really happening usually means hopping between campaign views, building custom reports, and exporting CSVs.

Understanding why performance changed, which campaigns are wasting budget, which keywords are raising your CPA, and what to do next: that analysis still takes time.

🔗 Connect your Google Ads account: https://onboard.windsor.ai/app/google_ads.

Once connected, you can ask ChatGPT questions that go far beyond your Google Ads dashboard:

• Which campaigns are spending the most but generating zero conversions this month?
• Where is my budget being lost to low-quality search terms?
• Which keywords with a Quality Score below 5 cost me the most?
• How does my ROAS compare across campaign types, Search, Shopping, Performance Max?
• What is my true ROAS by acquisition channel when Google Ads is blended with Meta and Shopify revenue?

Combining sources needs a paid plan. The free forever plan includes 1 data source.

And beyond analysis, the Windsor.ai plugin allows you to execute changes directly from ChatGPT. Find high CPA campaigns, wasted spend, or low-performing keywords, then pause or reallocate budget to top performers, without switching to Google Ads.

Steps to connect Google Ads to ChatGPT with the Windsor.ai plugin

The Google Ads to ChatGPT integration via Windsor takes two simple steps and requires no technical setup. The whole setup takes about a minute. For the short overview, see the Google Ads to ChatGPT integration page.

📖 Full documentation: windsor.ai/documentation/windsor-mcp/how-to-integrate-data-into-chatgpt/.

Prerequisites

Before connecting Google Ads to ChatGPT, make sure you have:

• An active Windsor.ai account. The free forever plan covers 1 user, 1 data source and 1 account with the last 30 days of data, and needs no credit card. Paid plans start at $19/month for 3 sources, billed annually (pricing).
• A Google Ads account (with at least Read-only access). Write actions need Standard or Admin access.
• An active ChatGPT account. We used ChatGPT’s Free plan for the steps on this page. In a Business, Enterprise or Edu workspace, an admin can turn plugins on or off.

Step 1. Connect Google Ads to Windsor

Go to onboard.windsor.ai/app/google_ads and connect Google Ads as your data source.

Sign in with the Google account that has access to your Ads account and approve the connection. Select the account(s) you want to pull data from. Connecting multiple accounts lets you perform cross-account analysis inside a single ChatGPT conversation. Tracking more than one account needs a paid plan. The free forever plan includes 1 account.

Windsor will pull your campaign data directly from the Google Ads API and normalize it in the background, making it structured and analysis-ready before it reaches ChatGPT.

Step 2. Add the Windsor.ai plugin in ChatGPT

Open the Windsor.ai plugin page in ChatGPT and select Install plugin. You can also click Plugins in the ChatGPT sidebar and search for Windsor.ai.

Complete the setup by granting ChatGPT access to your Windsor.ai account. To disconnect later, open Settings, then Plugins.

The Windsor.ai plugin is now connected. To confirm everything is working, start a new chat, click + and choose Windsor.ai (or type @Windsor), and run a quick test with the prompt:

List the data sources connected to my Windsor account.
Do you see Google Ads there?

Once ChatGPT recognizes your Google Ads data, you’re ready to start querying it in natural language. No ongoing maintenance is required.
Want the full walkthrough with real example prompts for pausing campaigns, shifting budget, changing bidding strategy, and managing Customer Match audiences? See ‘How to manage Google Ads campaigns from ChatGPT.’

The Google Ads connector has 34 write actions (write actions list): pause or enable a campaign, ad group, or ad; create a campaign, ad group, responsive search ad, or ad asset; rename a campaign or ad group, or update an ad group’s settings; set a campaign’s budget, bidding strategy, target CPA, target ROAS or CPC bid ceiling, and an ad group’s max CPC; set a campaign’s locations, languages and ad schedule; push, update or remove keywords, and push or remove negative keywords; and create, fill, rename, delete, attach or detach Customer Match audience lists.

Example: our top Google Ads campaigns by clicks in ChatGPT

We asked ChatGPT, through the Windsor.ai plugin: “For my Google Ads account windsor-main, which 5 campaigns got the most clicks in the last 30 days (6 October to 4 November 2026)? For each one, show only clicks, impressions and CTR.”

Our brand search campaign led with 3,660 clicks from 9,800 impressions, a CTR of 37.35%. The Performance Max campaign had the most impressions (179,115) and the lowest CTR of the five, 0.88%.

We checked the totals in a second question in the same chat (11 campaigns, 11,108 clicks, 378,728 impressions) and against the same data pulled from Windsor.ai directly.

Is there an official Google Ads app for ChatGPT?

No, as far as we could find. Google has no Google Ads plugin in ChatGPT’s Plugins directory. These are the routes that work, and when each fits:

• Windsor.ai plugin: reads your campaigns, keywords, search terms and Quality Scores when you ask, and runs the 34 Google Ads write actions after you confirm each change. It fits ongoing questions, changes from the chat, and Google Ads next to other sources.
• Google Ads MCP server: Google’s open-source Google Ads MCP server runs on your side, with a Google Cloud project and your own API credentials. It queries the account with GAQL and can create assets and manage labels; it has no tools to pause campaigns or change budgets and bids. In ChatGPT, a server like this is added as a custom MCP server from Plugins. Full support, including tools that change data, is limited to Business, Enterprise and Edu workspaces; Pro gets read and fetch tools only. It fits teams with developers who want GAQL access from the chat.
• Third-party plugins: other vendors publish Google Ads plugins for ChatGPT, for example Supermetrics.
• CSV export: download a Google Ads report and upload it to a chat. It fits a one-off question about one report.
• Google Sheets: send Google Ads data to a sheet on a schedule with Windsor.ai and point ChatGPT or a dashboard at that copy. It fits shared reports that keep their own history.

What Google Ads data Windsor sends to ChatGPT

Windsor connects to the Google Ads API and gives ChatGPT access to 2,542 Google Ads fields (495 metrics and 2,047 dimensions). View the full list of supported data fields.

The most commonly used fields in real performance analysis, and what ChatGPT can answer with them:

• Campaign performance: Cost, conversions, conversion value, ROAS, CPA, impressions, clicks, CTR: the core efficiency metrics at the campaign level
• Ad group & keyword data: Keyword-level CPA, CPC, Quality Score, match type, search term report
• Search terms: Actual queries triggering your ads: the dataset for negative keyword identification and match type analysis
• Quality Score components: Quality Score (1 to 10), Expected CTR rating, Ad Relevance rating, Landing Page Experience rating: the three components Google rates for each keyword
• Bidding & budget: Campaign budget, impression share lost to budget, impression share lost to rank, bidding strategy type: for budget allocation and bid strategy decisions
• Audience segments: In-market audiences, remarketing lists, customer match, similar audiences: performance and CPA by audience layer
• Device & time: Device split (mobile/desktop/tablet), hour of day, day of week: for dayparting and device bid adjustments
• Asset groups (PMax): Performance Max asset group performance, cost and conversions per asset group: the closest thing to transparency inside PMax
• Shopping & product data: Product ID, product title, product category: for Shopping and PMax product-level performance analysis
• Conversion actions: Conversion action name and count: breakdown of form fills vs. calls vs. purchases vs. other conversion types

Prompt ideas: What you can do with Google Ads data in ChatGPT

The use cases and prompts below cover the most common and valuable Google Ads analysis workflows.

To jumpstart, just replace campaign names, targets, and thresholds with the values from your own account.

🤖 For more prompt ideas across Google Ads and other data sources, visit windsor.ai/prompt-library/.

1. Finding budget waste: where spend isn’t delivering results

Budget waste in Google Ads is rarely visible at the campaign level; it hides in specific keywords, match types, and search terms that accumulate spend over weeks.

ChatGPT can surface it across your entire account with a single prompt.

Detect keywords with spend but zero conversions

Prompt:

Pull all keywords with more than $50 in spend and zero conversions in the last 30 days.

For each keyword, show:
- keyword_text, match_type, campaign_name, ad_group_name
- cost, impressions, clicks, ctr
- average_cpc, quality_score

Sort by cost descending (these are our clearest budget waste candidates).

For the top 20 by spend: are any currently in campaigns using Max Conversions bidding?
(If so, the algorithm may be bidding aggressively on terms it can't convert. Flag them for bid strategy review.)

Search term analysis: which queries trigger your ads

Prompt:

From the search term report for the last 30 days, show all search terms with more than $20 in spend.

For each term, show:
- search_term, campaign_name, ad_group_name, match_type
- cost, conversions, conversion_rate, cost_per_conversion

Group into three buckets:
1. High performers: conversion_rate above account average
2. Zero converters: spend > $20, zero conversions
3. Irrelevant terms: search terms that clearly don't match the advertised product or service (flag based on content mismatch)

For bucket 2 and 3, what negative keywords should be added to prevent future spend on these queries?

2. Evaluating Quality Score and CPC efficiency

Quality Score is Google’s 1 to 10 rating of how your ads and landing pages compare with other advertisers’ for each keyword. Google calls it a diagnostic tool, not an input in the ad auction, and builds it from three components: expected CTR, ad relevance and landing page experience (Google Ads Help).

ChatGPT can triage your entire keyword list by QS component and tell you where to focus.

Quality Score triage across the account

Prompt:

Pull all keywords with quality_score of 5 or below, with more than $30 in spend in the last 30 days.

For each keyword show:
- keyword_text, campaign_name, ad_group_name
- quality_score, search_predicted_ctr, creative_quality_score, post_click_quality_score
- cost, average_cpc

For the top 20 by spend:
- What is the most common failing component (expected CTR, ad relevance, or landing page experience)?
- Group by primary failure type
- If landing page experience is the common issue, list the affected campaigns and ad groups (this suggests a structural mismatch between ad copy and destination page)

3. Analyzing ROAS and campaign performance: finding what’s working

Understanding which campaigns are truly profitable goes beyond just looking at spend and conversions. True ROAS accounts for all revenue sources and accurately measures which campaigns are generating value.

ChatGPT can rank your campaigns by true ROAS and highlight where budget is most efficiently spent, so you know what’s working and what’s dragging down performance.

Campaign performance ranked by true ROAS

Prompt:

For all active campaigns in the last 30 days, show:
- campaign_name, campaign_type (Search, Shopping, PMax, Display, Video)
- cost, conversions, conversion_value
- roas (conversion_value / cost)
- cpa (cost / conversions)
- search_impression_share, search_budget_lost_impression_share, search_rank_lost_impression_share

Rank by ROAS descending.

Which campaigns are above your target ROAS but losing impression share due to budget? (These are our highest-priority candidates for budget reallocation)
Which campaigns are below target ROAS AND losing impression share to rank? (Those need both bid and quality work before more budget helps)

Performance Max transparency: what’s driving results inside your PMax campaigns

Prompt:

For all Performance Max campaigns, break down performance by asset group:
- asset_group_name, cost, conversions, conversion_value, roas

Which asset groups are below our target ROAS?
Which asset groups account for more than 50% of PMax spend?

Also pull top products within PMax campaigns (if Shopping feed is connected):
- product_item_id, product_title, cost, conversions

Flag any product that has spent more than $50 with zero conversions in the last 14 days (these products may need to be excluded from PMax or given dedicated Shopping campaigns with tighter controls.)

4. Estimating B2B and lead generation: connecting Google Ads to pipeline quality

For B2B teams, Google Ads conversions are form fills. B2B teams need to know which campaigns produce a qualified pipeline, and at what cost.

🔗 Connect Windsor with your CRM (HubSpot, Salesforce, etc.) data alongside Google Ads for full-funnel reporting.

Cost per qualified lead by campaign

Prompt:

Join Google Ads campaign data (campaign_name, cost, conversions) with CRM lead data using utm_campaign as the join key.

For each campaign, show:
- Total ad spend
- Total leads (Google Ads conversions)
- Leads that progressed to SQL stage (from CRM)
- SQL conversion rate
- Cost per SQL (spend/SQLs)
- Average deal value of opportunities from this campaign

Rank by cost per SQL ascending.

Which campaigns produce cheap leads that never qualify?
Which have fewer total conversions but a high SQL rate? (Our highest-quality traffic sources)

Combining sources needs a paid plan. The free forever plan includes 1 data source.

5. Agency use: cross-account analysis and client reporting

For agencies managing multiple Google Ads accounts, Windsor connects at the MCC level; all accounts are available in a single ChatGPT conversation for cross-account comparison and portfolio-level analysis. You can also filter by a specific account to generate separate reports. Tracking more than one account needs a paid plan. The free forever plan includes 1 account.

Cross-account budget pacing and anomaly detection

Prompt:

For each connected Google Ads account, pull campaign, budget_amount and cost by date for this month so far.

Calculate projected month-end spend at current run rate.

Flag any account where projected spend is more than 15% above or below the monthly budget.

Also flag any campaign where yesterday's cost was more than 3x its 7-day average daily spend (this usually signals a bidding anomaly or an accidental budget increase).

Client performance narrative for monthly reporting

Prompt:

Using Google Ads data for [client name] for this month vs. last month:

Compare: cost, impressions, clicks, ctr, conversions, cpa, conversion_value, roas.

Write a 3-paragraph client performance summary that:

1. Leads with the headline result: better or worse efficiency, and by how much
2. Explains the two biggest drivers of that change at campaign or ad group level
3. Gives one specific recommendation for next month with expected impact

Write for a non-technical client who cares about results.

No raw numbers in isolation: provide context.

Prompts that cover more than 30 days need a paid plan. The free forever plan includes the last 30 days.

Bonus: Blending Google Ads with other data sources in ChatGPT

Questions about whether ad spend brings revenue, how Google compares with other channels, and what customers do after they convert need data from outside Google Ads.

🔗 With Windsor, you can connect all your business and marketing tools (350+ sources supported), giving you a complete cross-channel overview in a single chat.

Here are some helpful analytical scenarios with multi-channel data you can try:

• Google Ads + Meta Ads: Compare efficiency across paid channels in a single conversation. Ask ChatGPT to show blended ROAS and CPA across Google and Meta, with Google’s impression share alongside, and recommend where to shift budget based on performance data from both platforms simultaneously.
• Google Ads + Shopify or WooCommerce: Google Ads reports conversion value based on what the pixel fires. Your store records actual completed orders. Connect both and ask ChatGPT to reconcile reported ROAS against actual net revenue after returns and failed payments, by campaign and by product category.
• Google Ads + GA4: Google Ads reports that someone converted. GA4 records what visitors did on your site after the click. Connect both to ask ChatGPT: which campaigns have strong Google Ads CTR but high GA4 bounce rate? Where is there a disconnect between the ad promise and the landing page experience?
• Google Ads + Salesforce or HubSpot: For B2B teams, the most important analysis is connecting Google Ads spend to the CRM pipeline. Ask ChatGPT to join campaign data with deal data using utm_campaign as the key, calculating cost per SQL, cost per closed deal, and which campaigns produce closed revenue.
• Google Ads + Google Sheets: Many teams track targets, budget caps, and account notes in spreadsheets. Connect your planning sheets alongside Google Ads data and ask ChatGPT to compare actual pacing against your planned budget allocation, or flag campaigns that have hit their monthly spend cap with days remaining in the period.

Combining sources needs a paid plan. The free forever plan includes 1 data source.

How the Windsor.ai plugin works with Google Ads and ChatGPT

Windsor connects to the Google Ads API and pulls structured campaign data, such as metrics, dimensions, and diagnostic fields, into your ChatGPT conversations through the Windsor.ai plugin.

This is one of the fastest and simplest ways to get your Google Ads data into ChatGPT. There is no export step and no scheduled sync to manage.

Windsor.ai pulls the data from the Google Ads API when you ask. The connection includes:

• Data when you ask. ChatGPT requests the fields a question needs from Windsor.ai, so you do not export a report first.
• 2,542 Google Ads fields. Windsor gives ChatGPT access to the full Google Ads data model: campaign structure, performance metrics, Quality Score diagnostics, audience data, asset group breakdowns, and search term reports.
• Cross-source blending. Windsor connects 350+ data sources. Google Ads data can be analysed alongside Meta Ads, Shopify, GA4, Salesforce, and more in a single ChatGPT conversation.
• MCC support. If you manage multiple Google Ads accounts through a Manager Account (MCC), Windsor connects at the MCC level, making cross-account analysis available in a single chat.
• Write actions. ChatGPT can make the changes in the write actions list in your Google Ads account, after you confirm each one.

If something does not work

• A Google Ads account is missing: connect again in Windsor.ai with a Google account that has access to it, and select it in the account list. Tracking more than one account needs a paid plan. The free forever plan includes 1 account.
• Google keeps asking for a passkey when you connect Google Ads: create one at g.co/passkeys and sign in again. A new passkey can take up to 7 days before Google accepts it. On a Google Workspace account, an admin can turn off the login challenge for your user for 10 minutes. During that window you can approve the passkey and connect. The Google Ads connector setup guide has the steps.
• ChatGPT answers without your Google Ads data: choose Windsor.ai under + or type @Windsor in that chat, then ask again.
• Windsor.ai shows “Disabled by admin”: your ChatGPT workspace admin has turned the plugin off. Ask them to enable it.
• A change is refused for lack of access: the Google account you connected has Read only access; switch to a login with Standard or Admin access.
• A max CPC change is refused: the ad group’s campaign uses Smart Bidding. Max CPC bids apply under Manual CPC, so switch the campaign’s bidding strategy first or keep the automated bids.
• Quality Score is empty for a keyword: Google leaves it blank when it does not have enough data, often for keywords with few impressions.

The Google Ads to ChatGPT integration page gives the short answers and a second example, our campaigns compared by cost per conversion.

Conclusion

With Windsor.ai installed, ChatGPT reads your Google Ads campaigns, keywords, search terms and Performance Max asset groups when you ask about them. It can also make changes through write actions, such as pausing a campaign, adding negative keywords or setting a campaign budget.

Add Meta Ads, GA4 or your CRM through the same connection to compare channels, or to see which campaigns brought closed deals, in the same chat.

A weekly Google Ads report can also run without a prompt. In a ChatGPT Business, Enterprise or Edu workspace, a workspace agent can send the weekly Google Ads report to your team on a schedule. Our Claude Cowork vs ChatGPT Work test ran the same scheduled report in both assistants, and both got every total right.

Companies on Gemini Enterprise can follow how to connect Google Ads to Gemini Enterprise.

🚀 Connect Google Ads to ChatGPT with the Windsor.ai plugin in about a minute. Start for free now → onboard.windsor.ai/app/google_ads.