Every Google Ads manager has faced the same challenge.
The account is full of data: spend, impressions, conversions, Quality Scores, search terms, device splits, and thousands 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really happening usually means hopping between campaign views, building custom reports, and exporting CSVs that are outdated before you even open them.
Understanding why performance changed, which campaigns are wasting budget, which keywords are quietly dragging down your CPA, and what to do next, that analysis still takes time that most teams don’t have.
With Windsor MCP, your Google Ads data streams live into ChatGPT in under a minute, unlocking deep AI insights and actionable optimization tips instantly. All it takes is connecting your Google Ads account to Windsor.ai and installing the Windsor.ai app for ChatGPT.
🔗 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 have a Quality Score below 5 and are costing me the most in inflated CPCs?
- 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?
And beyond analysis, Windsor MCP 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 instantly, without switching to Google Ads Manager.
Ready to start automating your Google Ads analysis in ChatGPT? Let’s walk through the steps to build this integration.
Steps to connect Google Ads to ChatGPT with Windsor MCP
The Google Ads to ChatGPT integration via Windsor takes two simple steps and requires no technical setup.
📖 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 (free or paid plan)
- A Google Ads account (with at least Read-only access)
- An active ChatGPT account
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.

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. Install the Windsor.ai app for ChatGPT
Open the Windsor.ai app page in ChatGPT and click Connect.

Complete the setup by granting ChatGPT access to your Windsor.ai account.
Windsor MCP is now connected. To confirm everything is working, start a new chat 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.’
Currently available write actions: pause or enable a campaign, ad group, or ad; set a campaign’s budget or bidding strategy; push, update, or remove keywords and negative keywords; manage Customer Match audience lists; and create a campaign, ad group, or responsive search ad.
What Google Ads data Windsor sends to ChatGPT
Windsor connects to the Google Ads API and gives ChatGPT access to over 2,300 Google Ads metrics and dimensions. View the full list of supported data fields.
The most commonly used fields in real performance analysis:
| Google Ads data | What ChatGPT can answer with it |
| 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: the granular layer where most budget waste lives |
| Search terms | Actual queries triggering your ads: the most actionable dataset for negative keyword identification and match type analysis |
| Quality Score components | Quality Score (1–10), Expected CTR rating, Ad Relevance rating, Landing Page Experience rating: the three levers that control CPCs |
| 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 quietly.
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 for bid strategy review.)
Search term analysis: what queries are actually triggering 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 has a direct, compounding effect on CPCs. A keyword with QS 4 can cost up to 150% more per click than the same keyword with QS 8.
ChatGPT can triage your entire keyword list by QS component and tell you exactly 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, expected_ctr_rating, ad_relevance_rating, landing_page_experience_rating - cost, average_cpc For the top 20 by spend: - What is the most common failing component (CTR, relevance, or landing page)? - 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) - Estimate the CPC reduction if QS improved to 7 for each keyword (QS 7 = ~15% CPC reduction vs. QS 5)
3. Analyzing ROAS and campaign performance: finding what’s actually 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 exactly 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) - impression_share, impression_share_lost_budget, impression_share_lost_rank 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 product IDs within PMax campaigns (if Shopping feed is connected): - product_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 fillsm not revenue. The real question is 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)
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.
Cross-account budget pacing and anomaly detection
Prompt:
For each connected Google Ads account, pull: - campaign_name, monthly_budget, cost_yesterday, cost_month_to_date, days_remaining_in_month 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, not metrics. No raw numbers in isolation — provide context.
Bonus: Blending Google Ads with other data sources in ChatGPT
Google Ads tells you what happened inside the platform. But the most valuable questions, such as whether ad spend is generating actual revenue, how Google compares to other channels, and what customers do after they convert, require data that lives 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, CPA, and impression share across Google and Meta, 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, by product category, and by customer segment.
- Google Ads + GA4: Google Ads tells you someone converted. GA4 tells you what they did before that. 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 revenue, not just leads.
- 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.
How Windsor MCP 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 direct, fully automated connector.
This is one of the fastest and simplest ways to get your Google Ads data into ChatGPT. There is no export step, no scheduled sync to manage, and no stale data.
Every time you ask a question, Windsor retrieves the current state of your account thanks to these features:
- Live campaign data. Windsor pulls from the Google Ads API in real time. The data ChatGPT analyzes reflects your account as it stands right now, not a snapshot from last week’s export.
- 2,350+ 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-enabled actions. Execute approved changes directly in your Google Ads account: create campaigns, ad groups, and ads; pause or enable ad groups and ads; set the bid strategy or max CPC; and push new negative keywords, all based on your prompts and insights.
Conclusion
In Google Ads, the difference between a high-performing account and one that wastes money is often in the data, but finding it takes time. Natively, the platform is designed for tracking activity, not giving insights.
With Windsor MCP, your Google Ads data flows live into ChatGPT, so you can instantly see which campaigns are winning, get recommendations for fixing issues like wasted budget, low Quality Score keywords, irrelevant search terms, or underperforming PMax asset groups, and make changes, all without hours of manual work.
And when you need answers beyond Google Ads, like your true ROAS, including returns, which campaigns generate real deals versus just leads, or how Google performs compared to Meta this quarter, Windsor connects all your tools, giving you a complete cross-channel analysis in one chat.
🚀 Connect Google Ads to ChatGPT with Windsor MCP in less than a minute. Start for free now → onboard.windsor.ai/app/google_ads.
Available actions
Push Negative Keywords Add negative keywords to a Google Ads campaign or ad group. Set level to 'campaign' (with campaign_id) to block the terms across the whole campaign, or 'ad_group' (with ad_group_id) to block them only in that ad group. Each keyword has a match type (BROAD, PHRASE or EXACT, defaulting to BROAD). Keywords are pushed in one request; any that Google rejects (for example a duplicate, or a negative that already exists on the target) are skipped and reported, and the remaining valid keywords are still added. To read back the pushed negative keywords, query get_data with the campaign_criterion_keyword_text, campaign_criterion_keyword_match_type and campaign_criterion_negative fields (campaign level) or the ad_group_criterion_keyword_text, ad_group_criterion_keyword_match_type and ad_group_criterion_negative fields (ad group level).
Add negative keywords to a Google Ads campaign or ad group. Set level to 'campaign' (with campaign_id) to block the terms across the whole campaign, or 'ad_group' (with ad_group_id) to block them only in that ad group. Each keyword has a match type (BROAD, PHRASE or EXACT, defaulting to BROAD). Keywords are pushed in one request; any that Google rejects (for example a duplicate, or a negative that already exists on the target) are skipped and reported, and the remaining valid keywords are still added. To read back the pushed negative keywords, query get_data with the campaign_criterion_keyword_text, campaign_criterion_keyword_match_type and campaign_criterion_negative fields (campaign level) or the ad_group_criterion_keyword_text, ad_group_criterion_keyword_match_type and ad_group_criterion_negative fields (ad group level).
Push Keywords Add positive keywords to a Google Ads ad group (positive keywords live on the ad group, not the campaign). Each keyword has a match type (BROAD, PHRASE or EXACT, defaulting to BROAD) and an optional max CPC bid in micros. Status ('enabled' or 'paused') defaults to 'enabled' for the whole batch and can be overridden per keyword. Keywords are pushed in one request; any that Google rejects (for example a duplicate that already exists in the ad group) are skipped and reported, and the remaining valid keywords are still added.
Add positive keywords to a Google Ads ad group (positive keywords live on the ad group, not the campaign). Each keyword has a match type (BROAD, PHRASE or EXACT, defaulting to BROAD) and an optional max CPC bid in micros. Status ('enabled' or 'paused') defaults to 'enabled' for the whole batch and can be overridden per keyword. Keywords are pushed in one request; any that Google rejects (for example a duplicate that already exists in the ad group) are skipped and reported, and the remaining valid keywords are still added.
Create Responsive Search Ad Create a responsive search ad inside an existing Search ad group. Provide 3-15 headlines (max 30 characters each) and 2-4 descriptions (max 90 characters each); Google mixes them into ad combinations. final_url is the landing page. The optional path1 and path2 are display-URL segments (max 15 characters each; path2 requires path1). The ad is created paused by default so it does not start serving immediately.
Create a responsive search ad inside an existing Search ad group. Provide 3-15 headlines (max 30 characters each) and 2-4 descriptions (max 90 characters each); Google mixes them into ad combinations. final_url is the landing page. The optional path1 and path2 are display-URL segments (max 15 characters each; path2 requires path1). The ad is created paused by default so it does not start serving immediately.
Create Ad Group Create a new ad group inside an existing Search or Display campaign. The ad group type is derived from the campaign's channel type. The optional cpc_bid_micros sets the default max CPC bid in micros of the account currency (1,000,000 micros = 1 unit). The ad group is created paused by default.
Create a new ad group inside an existing Search or Display campaign. The ad group type is derived from the campaign's channel type. The optional cpc_bid_micros sets the default max CPC bid in micros of the account currency (1,000,000 micros = 1 unit). The ad group is created paused by default.
Create Campaign Create a new Search or Display Google Ads campaign together with its daily budget in a single atomic request. The budget amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 50_000_000 means 50.00). The campaign is created paused by default so it does not start serving immediately. It is declared as not containing EU political advertising unless contains_eu_political_advertising is set to true. The maximize_conversions bidding strategy requires conversion tracking to be enabled on the account. Add ad groups to it with the create_ad_group action.
Create a new Search or Display Google Ads campaign together with its daily budget in a single atomic request. The budget amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 50_000_000 means 50.00). The campaign is created paused by default so it does not start serving immediately. It is declared as not containing EU political advertising unless contains_eu_political_advertising is set to true. The maximize_conversions bidding strategy requires conversion tracking to be enabled on the account. Add ad groups to it with the create_ad_group action.
Set Campaign Budget Set the daily or total (lifetime) budget of a Google Ads campaign. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 50_000_000 means 50.00). A lifetime budget only applies to campaigns with a fixed end date. If the budget is also used by other campaigns, the change is rejected unless apply_to_shared_budget is set to true, because it would affect every campaign that shares that budget.
Set the daily or total (lifetime) budget of a Google Ads campaign. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 50_000_000 means 50.00). A lifetime budget only applies to campaigns with a fixed end date. If the budget is also used by other campaigns, the change is rejected unless apply_to_shared_budget is set to true, because it would affect every campaign that shares that budget.
Create Ad Asset Create a Google Ads ad extension asset and link it to a campaign or ad group in one atomic request. Supported asset_type values: 'sitelink' (extra links under the ad; needs link_text and final_url, with optional description_1 and description_2), 'callout' (short non-clickable highlight; needs callout_text), 'structured_snippet' (a header plus 3-10 values; needs header and values) and 'call' (a phone number; needs phone_number and country_code). Set level to 'campaign' (with campaign_id) to apply the extension across the whole campaign, or 'ad_group' (with ad_group_id) to apply it to one ad group only. Each call creates one asset; call the action again for each additional extension. New call and sitelink assets go through Google's policy review before they start showing.
Create a Google Ads ad extension asset and link it to a campaign or ad group in one atomic request. Supported asset_type values: 'sitelink' (extra links under the ad; needs link_text and final_url, with optional description_1 and description_2), 'callout' (short non-clickable highlight; needs callout_text), 'structured_snippet' (a header plus 3-10 values; needs header and values) and 'call' (a phone number; needs phone_number and country_code). Set level to 'campaign' (with campaign_id) to apply the extension across the whole campaign, or 'ad_group' (with ad_group_id) to apply it to one ad group only. Each call creates one asset; call the action again for each additional extension. New call and sitelink assets go through Google's policy review before they start showing.
Enable Campaign Enable a paused Google Ads campaign. Reversible via pause_campaign.
Enable a paused Google Ads campaign. Reversible via pause_campaign.
Pause Campaign Pause an active Google Ads campaign. Reversible via enable_campaign.
Pause an active Google Ads campaign. Reversible via enable_campaign.
Enable Ad Enable a paused ad within an ad group. Identify the ad by its ad group id and ad id; the create ad actions return these joined as '<ad_group_id>~<ad_id>'. Reversible via pause_ad.
Enable a paused ad within an ad group. Identify the ad by its ad group id and ad id; the create ad actions return these joined as '<ad_group_id>~<ad_id>'. Reversible via pause_ad.
Pause Ad Pause an active ad within an ad group. Identify the ad by its ad group id and ad id; the create ad actions return these joined as '<ad_group_id>~<ad_id>'. Reversible via enable_ad.
Pause an active ad within an ad group. Identify the ad by its ad group id and ad id; the create ad actions return these joined as '<ad_group_id>~<ad_id>'. Reversible via enable_ad.
Update Keywords Update existing positive keywords in a Google Ads ad group, identified by ad group criterion id (from the get_data 'keyword_criterion_id' field). Each update can change the keyword's status ('enabled' or 'paused') and/or its max CPC bid in micros; a keyword's text and match type cannot be changed (remove and re-add to change those). Updates are applied in one request; any that Google rejects (for example an unknown criterion id) are skipped and reported, and the remaining valid updates are still applied.
Update existing positive keywords in a Google Ads ad group, identified by ad group criterion id (from the get_data 'keyword_criterion_id' field). Each update can change the keyword's status ('enabled' or 'paused') and/or its max CPC bid in micros; a keyword's text and match type cannot be changed (remove and re-add to change those). Updates are applied in one request; any that Google rejects (for example an unknown criterion id) are skipped and reported, and the remaining valid updates are still applied.
Enable Ad Group Enable a paused Google Ads ad group. Reversible via pause_ad_group.
Enable a paused Google Ads ad group. Reversible via pause_ad_group.
Set Campaign Bidding Strategy Switch a Google Ads campaign to a different standard (campaign-level) bidding strategy. Supported strategies: manual_cpc, target_cpa, target_roas, maximize_conversions, maximize_conversion_value, maximize_clicks, target_spend. maximize_clicks and target_spend are aliases for the same strategy (the former is the Google Ads UI label, the latter the API and get_data field naming). target_cpa_micros is required for target_cpa and maximize_conversions (1,000,000 micros = 1 unit of the account currency). target_roas is required for target_roas and maximize_conversion_value (expressed as a ratio, e.g. 4.0 means 400% return on ad spend). cpc_bid_ceiling_micros is required for maximize_clicks and target_spend; it is not accepted for other strategies because the API only allows a campaign-level ceiling on Target Spend (other strategies require a portfolio bidding strategy). manual_cpc takes no additional parameters.
Switch a Google Ads campaign to a different standard (campaign-level) bidding strategy. Supported strategies: manual_cpc, target_cpa, target_roas, maximize_conversions, maximize_conversion_value, maximize_clicks, target_spend. maximize_clicks and target_spend are aliases for the same strategy (the former is the Google Ads UI label, the latter the API and get_data field naming). target_cpa_micros is required for target_cpa and maximize_conversions (1,000,000 micros = 1 unit of the account currency). target_roas is required for target_roas and maximize_conversion_value (expressed as a ratio, e.g. 4.0 means 400% return on ad spend). cpc_bid_ceiling_micros is required for maximize_clicks and target_spend; it is not accepted for other strategies because the API only allows a campaign-level ceiling on Target Spend (other strategies require a portfolio bidding strategy). manual_cpc takes no additional parameters.
Set CPC Bid Ceiling Set the maximum CPC bid the strategy is allowed to place on a Google Ads campaign. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 2_500_000 means 2.50). Only applies to campaigns using Target Spend / Maximize Clicks bidding. Other strategies (Target CPA, Target ROAS, Maximize Conversions, Maximize Conversion Value) treat ceiling as portfolio-only per the Google Ads API, so those cannot be adjusted via this action.
Set the maximum CPC bid the strategy is allowed to place on a Google Ads campaign. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 2_500_000 means 2.50). Only applies to campaigns using Target Spend / Maximize Clicks bidding. Other strategies (Target CPA, Target ROAS, Maximize Conversions, Maximize Conversion Value) treat ceiling as portfolio-only per the Google Ads API, so those cannot be adjusted via this action.
Remove Keywords Remove positive keywords from a Google Ads ad group, identified by ad group criterion id (from the get_data 'keyword_criterion_id' field). Removals are applied in one request; any that Google rejects (for example an unknown criterion id) are skipped and reported, and the remaining valid removals still apply.
Remove positive keywords from a Google Ads ad group, identified by ad group criterion id (from the get_data 'keyword_criterion_id' field). Removals are applied in one request; any that Google rejects (for example an unknown criterion id) are skipped and reported, and the remaining valid removals still apply.
Pause Ad Group Pause an active Google Ads ad group. Reversible via enable_ad_group.
Pause an active Google Ads ad group. Reversible via enable_ad_group.
Set Campaign Language Targeting Set the languages a Google Ads campaign targets, so its ads only reach users Google associates with one of them. Each language is given as its Google Ads language code, spelled exactly as the languages parameter enumerates it (e.g. en, de, es, zh_CN). Regional variants such as en_US and en_GB are not among them and cannot be targeted, only the base en. Google's table of these codes with the language each one names is at https://developers.google.com/google-ads/api/data/codes-formats#languages. This replaces the whole language targeting — every language to target must be listed in one call, because the languages not listed are removed. Languages already targeted are left in place, so their criterion ids do not change. A campaign with no language criteria targets all languages; pass an empty list to clear the targeting and return it to that state. Language targeting cannot be negative: there is no way to exclude a single language, only to list the ones to target. Google rejects the change for campaign types that do not support language criteria (e.g. Smart campaigns).
Set the languages a Google Ads campaign targets, so its ads only reach users Google associates with one of them. Each language is given as its Google Ads language code, spelled exactly as the languages parameter enumerates it (e.g. en, de, es, zh_CN). Regional variants such as en_US and en_GB are not among them and cannot be targeted, only the base en. Google's table of these codes with the language each one names is at https://developers.google.com/google-ads/api/data/codes-formats#languages. This replaces the whole language targeting — every language to target must be listed in one call, because the languages not listed are removed. Languages already targeted are left in place, so their criterion ids do not change. A campaign with no language criteria targets all languages; pass an empty list to clear the targeting and return it to that state. Language targeting cannot be negative: there is no way to exclude a single language, only to list the ones to target. Google rejects the change for campaign types that do not support language criteria (e.g. Smart campaigns).
Remove Negative Keywords Remove negative keywords from a Google Ads campaign or ad group, identified by criterion id. Set level to 'campaign' (with campaign_id) or 'ad_group' (with ad_group_id) to match where the negative keywords are attached. Removals are applied in one request; any that Google rejects (for example an unknown criterion id) are skipped and reported, and the remaining valid removals still apply. Reversible via push_negative_keywords by re-adding the keyword text and match type.
Remove negative keywords from a Google Ads campaign or ad group, identified by criterion id. Set level to 'campaign' (with campaign_id) or 'ad_group' (with ad_group_id) to match where the negative keywords are attached. Removals are applied in one request; any that Google rejects (for example an unknown criterion id) are skipped and reported, and the remaining valid removals still apply. Reversible via push_negative_keywords by re-adding the keyword text and match type.
Set Max CPC Bid Set the default Max CPC bid on a Google Ads ad group. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 2_500_000 means 2.50). Manual CPC bids live on the ad group, not the campaign, so the parent campaign must use the Manual CPC bidding strategy; switch it first via set_campaign_bidding_strategy if it uses Smart Bidding.
Set the default Max CPC bid on a Google Ads ad group. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 2_500_000 means 2.50). Manual CPC bids live on the ad group, not the campaign, so the parent campaign must use the Manual CPC bidding strategy; switch it first via set_campaign_bidding_strategy if it uses Smart Bidding.
Set Ad Schedule Set the ad schedule (dayparting) of a Google Ads campaign: the days and times of day its ads are allowed to serve, each with an optional bid adjustment. This replaces the campaign's whole schedule — every window it should serve in must be listed in one call, because the windows not listed are removed. Times are in the campaign's account time zone. Hours are 24-hour, minutes are limited to 0, 15, 30 and 45, and each window is exclusive of its end time, so 09:00-17:00 serves up to but not including 17:00; use end_hour 24 for the end of the day. Windows for the same day must not overlap, and a day takes at most six of them. A campaign with no ad schedule serves at all times; pass an empty schedule list to clear the schedule and return it to that state. To read the current schedule, query get_data with the campaign_criterion_ad_schedule_day_of_week, campaign_criterion_ad_schedule_start_hour, campaign_criterion_ad_schedule_start_minute, campaign_criterion_ad_schedule_end_hour and campaign_criterion_ad_schedule_end_minute fields.
Set the ad schedule (dayparting) of a Google Ads campaign: the days and times of day its ads are allowed to serve, each with an optional bid adjustment. This replaces the campaign's whole schedule — every window it should serve in must be listed in one call, because the windows not listed are removed. Times are in the campaign's account time zone. Hours are 24-hour, minutes are limited to 0, 15, 30 and 45, and each window is exclusive of its end time, so 09:00-17:00 serves up to but not including 17:00; use end_hour 24 for the end of the day. Windows for the same day must not overlap, and a day takes at most six of them. A campaign with no ad schedule serves at all times; pass an empty schedule list to clear the schedule and return it to that state. To read the current schedule, query get_data with the campaign_criterion_ad_schedule_day_of_week, campaign_criterion_ad_schedule_start_hour, campaign_criterion_ad_schedule_start_minute, campaign_criterion_ad_schedule_end_hour and campaign_criterion_ad_schedule_end_minute fields.
Set Target CPA Set the target cost per acquisition on a Google Ads campaign already using Target CPA or Maximize Conversions bidding. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 25_000_000 means 25.00). To switch a campaign to Target CPA bidding, use set_campaign_bidding_strategy with strategy='target_cpa' instead.
Set the target cost per acquisition on a Google Ads campaign already using Target CPA or Maximize Conversions bidding. The amount is in micros of the account currency (1,000,000 micros = 1 unit, e.g. 25_000_000 means 25.00). To switch a campaign to Target CPA bidding, use set_campaign_bidding_strategy with strategy='target_cpa' instead.
Rename Campaign Change the name of an existing Google Ads campaign. Only the name changes: budget, bidding, targeting, status, ad groups and ads are all left as they are, and historical reporting stays attached to the campaign because reports key on the campaign id. Campaign names must be unique within the account, so renaming to a name another campaign already uses is rejected. To rename an ad group instead, use the rename_ad_group action.
Change the name of an existing Google Ads campaign. Only the name changes: budget, bidding, targeting, status, ad groups and ads are all left as they are, and historical reporting stays attached to the campaign because reports key on the campaign id. Campaign names must be unique within the account, so renaming to a name another campaign already uses is rejected. To rename an ad group instead, use the rename_ad_group action.
Rename Ad Group Change the name of an existing Google Ads ad group. Only the name changes: bids, targeting, status, keywords and ads are all left as they are, and historical reporting stays attached to the ad group because reports key on the ad group id. Ad group names must be unique within their campaign, so renaming to a name another ad group in the same campaign already uses is rejected. To rename a campaign instead, use the rename_campaign action.
Change the name of an existing Google Ads ad group. Only the name changes: bids, targeting, status, keywords and ads are all left as they are, and historical reporting stays attached to the ad group because reports key on the ad group id. Ad group names must be unique within their campaign, so renaming to a name another ad group in the same campaign already uses is rejected. To rename a campaign instead, use the rename_campaign action.
Set Target ROAS Set the target return on ad spend on a Google Ads campaign already using Target ROAS or Maximize Conversion Value bidding. The value is a ratio of conversion value to spend, not a percentage (e.g. 4.0 means 400% ROAS). To switch a campaign to Target ROAS bidding, use set_campaign_bidding_strategy with strategy='target_roas' instead.
Set the target return on ad spend on a Google Ads campaign already using Target ROAS or Maximize Conversion Value bidding. The value is a ratio of conversion value to spend, not a percentage (e.g. 4.0 means 400% ROAS). To switch a campaign to Target ROAS bidding, use set_campaign_bidding_strategy with strategy='target_roas' instead.
Create Customer Match List Create an empty Google Ads Customer Match user list (CRM-based, with CONTACT_INFO as the upload key type — email, phone, or mailing address) that can then be populated with upload_customer_match_list. Returns the numeric user_list_id to pass to that action. membership_life_span_days controls how long an uploaded member remains in the list before Google auto-removes it: 1-540 for a fixed lifespan, or 10000 as Google's sentinel value for 'no expiration'. The account must be enrolled in Customer Match; if it is not, creation fails with a policy error and no list is created.
Create an empty Google Ads Customer Match user list (CRM-based, with CONTACT_INFO as the upload key type — email, phone, or mailing address) that can then be populated with upload_customer_match_list. Returns the numeric user_list_id to pass to that action. membership_life_span_days controls how long an uploaded member remains in the list before Google auto-removes it: 1-540 for a fixed lifespan, or 10000 as Google's sentinel value for 'no expiration'. The account must be enrolled in Customer Match; if it is not, creation fails with a policy error and no list is created.
Attach User List to Ad Group Attach a Customer Match user list to a Google Ads ad group so the ad group can target or exclude users in that list. Pass exclude=false (default) to add the list as a positive audience criterion, or exclude=true to add it as a negative criterion (never show ads to users in the list — commonly used to suppress existing customers). This action is idempotent: attaching a list that is already attached with the same exclude value is a no-op and returns the existing criterion. Attaching with a different exclude value than the current attachment returns an error — detach first, then re-attach. targeting_mode controls whether the list narrows delivery for the ad group ('targeting') or is only used for reporting and bidding ('observation'). Passing 'targeting' or 'observation' updates the ad group's AUDIENCE target_restriction; other target_restrictions (e.g. DEMOGRAPHIC) are preserved. Passing 'unchanged' (default) leaves targeting_setting untouched; note the setting is ad-group-wide, so changing it affects ALL audience criteria on the ad group, not just the list being attached. Negative (exclude=true) criteria always apply regardless of targeting_setting. Scope: this action is for AD-GROUP-level audiences only; campaign-level audience criteria and Performance Max audience signals use different APIs and are not supported here.
Attach a Customer Match user list to a Google Ads ad group so the ad group can target or exclude users in that list. Pass exclude=false (default) to add the list as a positive audience criterion, or exclude=true to add it as a negative criterion (never show ads to users in the list — commonly used to suppress existing customers). This action is idempotent: attaching a list that is already attached with the same exclude value is a no-op and returns the existing criterion. Attaching with a different exclude value than the current attachment returns an error — detach first, then re-attach. targeting_mode controls whether the list narrows delivery for the ad group ('targeting') or is only used for reporting and bidding ('observation'). Passing 'targeting' or 'observation' updates the ad group's AUDIENCE target_restriction; other target_restrictions (e.g. DEMOGRAPHIC) are preserved. Passing 'unchanged' (default) leaves targeting_setting untouched; note the setting is ad-group-wide, so changing it affects ALL audience criteria on the ad group, not just the list being attached. Negative (exclude=true) criteria always apply regardless of targeting_setting. Scope: this action is for AD-GROUP-level audiences only; campaign-level audience criteria and Performance Max audience signals use different APIs and are not supported here.
Upload Customer Match List Upload contact records (email, phone, or mailing address) to a Google Ads Customer Match user list for audience targeting. The user list must already exist in Google Ads; pass its numeric id as user_list_id. If you do not have one yet, call create_customer_match_list first to create an empty list and get its id. Uploads go through Google's Data Manager API (https://developers.google.com/data-manager/api/devguides/audiences/google-ads/customer-match), which requires the Google Ads account to have been authorised with the ``https://www.googleapis.com/auth/datamanager`` OAuth scope; the action returns a clear error if the scope is missing. This action is ASYNCHRONOUS: it returns immediately with a request_id once Google has accepted the request. Google then matches the records against signed-in users in the background — this typically takes a few minutes but can take up to 24 hours for large lists. The user list is NOT usable for targeting until matching completes. To check progress, call get_customer_match_upload_status with the returned request_id. Poll no more than once every 60 seconds. The job is done when status is SUCCESS, PARTIAL_SUCCESS, or FAILED; while it is PROCESSING keep polling. PII HANDLING: emails, phone numbers, and names are normalized and SHA-256 hashed by this action before being sent to Google — do NOT pre-hash them; pass raw values. operation_type=add appends records to the list; operation_type=remove removes matching records (use for compliance requests such as GDPR erasure). At most 10,000 records per call; split larger uploads into multiple calls.
Upload contact records (email, phone, or mailing address) to a Google Ads Customer Match user list for audience targeting. The user list must already exist in Google Ads; pass its numeric id as user_list_id. If you do not have one yet, call create_customer_match_list first to create an empty list and get its id. Uploads go through Google's Data Manager API (https://developers.google.com/data-manager/api/devguides/audiences/google-ads/customer-match), which requires the Google Ads account to have been authorised with the ``https://www.googleapis.com/auth/datamanager`` OAuth scope; the action returns a clear error if the scope is missing. This action is ASYNCHRONOUS: it returns immediately with a request_id once Google has accepted the request. Google then matches the records against signed-in users in the background — this typically takes a few minutes but can take up to 24 hours for large lists. The user list is NOT usable for targeting until matching completes. To check progress, call get_customer_match_upload_status with the returned request_id. Poll no more than once every 60 seconds. The job is done when status is SUCCESS, PARTIAL_SUCCESS, or FAILED; while it is PROCESSING keep polling. PII HANDLING: emails, phone numbers, and names are normalized and SHA-256 hashed by this action before being sent to Google — do NOT pre-hash them; pass raw values. operation_type=add appends records to the list; operation_type=remove removes matching records (use for compliance requests such as GDPR erasure). At most 10,000 records per call; split larger uploads into multiple calls.
Delete Customer Match List Permanently delete a Google Ads Customer Match user list. Detaches the list from any ad groups or campaigns currently using it as an audience or exclusion criterion and removes all uploaded members. Historical performance data for the list is retained in reports. Cannot be undone; to recover, create a new list and re-upload members. Use with care — deleting a list that is actively targeting a live campaign changes delivery immediately.
Permanently delete a Google Ads Customer Match user list. Detaches the list from any ad groups or campaigns currently using it as an audience or exclusion criterion and removes all uploaded members. Historical performance data for the list is retained in reports. Cannot be undone; to recover, create a new list and re-upload members. Use with care — deleting a list that is actively targeting a live campaign changes delivery immediately.
Detach User List from Ad Group Remove a Customer Match user list criterion (audience or exclusion) from a Google Ads ad group. Reverses attach_user_list_to_ad_group. Removes both positive (audience) and negative (exclusion) criteria — pass the same ad_group_id and user_list_id used when attaching. Removing a user list stops any targeting or reporting effect it had on future traffic; historical performance data remains in reports. If the list is not currently attached to the ad group, the action returns an error and no changes are made.
Remove a Customer Match user list criterion (audience or exclusion) from a Google Ads ad group. Reverses attach_user_list_to_ad_group. Removes both positive (audience) and negative (exclusion) criteria — pass the same ad_group_id and user_list_id used when attaching. Removing a user list stops any targeting or reporting effect it had on future traffic; historical performance data remains in reports. If the list is not currently attached to the ad group, the action returns an error and no changes are made.
Get Customer Match Upload Status Check the status of a Customer Match upload request started by upload_customer_match_list. Pass the request_id that action returned. Status is one of PROCESSING, SUCCESS, PARTIAL_SUCCESS, FAILED. PROCESSING means matching is still in progress — poll again in at least 60 seconds. Do NOT poll more frequently; quota is shared with data reads. If still PROCESSING after 24 hours, treat as stuck and contact support. SUCCESS means every uploaded record was accepted; the response includes the record count and match_rate_range, a coarse bucket indicating what fraction of uploaded records were matched to signed-in Google users. A low match rate usually reflects source data quality (invalid or non-Google-account emails/phones), not upload failure. For operation_type=remove requests, SUCCESS means the matching records were removed from the list. PARTIAL_SUCCESS means some records were accepted and others were rejected; the response includes the same counts plus the error message describing which records failed. The list still receives the accepted records. FAILED means the entire request failed with no records ingested; the response includes the reason. Failed requests cannot be resumed; start a new upload after fixing the issue.
Check the status of a Customer Match upload request started by upload_customer_match_list. Pass the request_id that action returned. Status is one of PROCESSING, SUCCESS, PARTIAL_SUCCESS, FAILED. PROCESSING means matching is still in progress — poll again in at least 60 seconds. Do NOT poll more frequently; quota is shared with data reads. If still PROCESSING after 24 hours, treat as stuck and contact support. SUCCESS means every uploaded record was accepted; the response includes the record count and match_rate_range, a coarse bucket indicating what fraction of uploaded records were matched to signed-in Google users. A low match rate usually reflects source data quality (invalid or non-Google-account emails/phones), not upload failure. For operation_type=remove requests, SUCCESS means the matching records were removed from the list. PARTIAL_SUCCESS means some records were accepted and others were rejected; the response includes the same counts plus the error message describing which records failed. The list still receives the accepted records. FAILED means the entire request failed with no records ingested; the response includes the reason. Failed requests cannot be resumed; start a new upload after fixing the issue.
Update Ad Group Partially update an existing Google Ads ad group. Provide only the fields to change; at least one besides ad_group_id is required and omitted fields are left untouched. status is 'enabled' or 'paused'; ad_rotation_mode is 'optimize' or 'rotate_forever'. cpc_bid_micros is the default max CPC bid in micros of the account currency (1,000,000 micros = 1 unit, e.g. 2_500_000 means 2.50) and only applies when the parent campaign uses Manual CPC; switch it first via set_campaign_bidding_strategy if it uses Smart Bidding. Budget, target CPA and CPM bids are not editable here. This is a value change with no built-in undo: to reverse it read the current values first (e.g. GAQL 'SELECT ad_group.name, ad_group.status, ad_group.cpc_bid_micros, ad_group.ad_rotation_mode FROM ad_group WHERE ad_group.id = <id>') then call update_ad_group again with the previous values. The single-purpose rename_ad_group, pause_ad_group/enable_ad_group and set_max_cpc actions still exist; prefer them for one change and use update_ad_group to change several fields at once.
Partially update an existing Google Ads ad group. Provide only the fields to change; at least one besides ad_group_id is required and omitted fields are left untouched. status is 'enabled' or 'paused'; ad_rotation_mode is 'optimize' or 'rotate_forever'. cpc_bid_micros is the default max CPC bid in micros of the account currency (1,000,000 micros = 1 unit, e.g. 2_500_000 means 2.50) and only applies when the parent campaign uses Manual CPC; switch it first via set_campaign_bidding_strategy if it uses Smart Bidding. Budget, target CPA and CPM bids are not editable here. This is a value change with no built-in undo: to reverse it read the current values first (e.g. GAQL 'SELECT ad_group.name, ad_group.status, ad_group.cpc_bid_micros, ad_group.ad_rotation_mode FROM ad_group WHERE ad_group.id = <id>') then call update_ad_group again with the previous values. The single-purpose rename_ad_group, pause_ad_group/enable_ad_group and set_max_cpc actions still exist; prefer them for one change and use update_ad_group to change several fields at once.
Rename Customer Match List Change the display name of a Google Ads Customer Match user list. Does not affect membership, campaign attachments, targeting behaviour, or the list's configuration — only the name shown in the Google Ads UI and reports. Names must be unique per account; Google returns an error if another list already uses the requested name.
Change the display name of a Google Ads Customer Match user list. Does not affect membership, campaign attachments, targeting behaviour, or the list's configuration — only the name shown in the Google Ads UI and reports. Names must be unique per account; Google returns an error if another list already uses the requested name.
Set Campaign Geo Targeting Set the geographic targeting of a Google Ads campaign: the countries, regions or cities its ads may serve in, plus optional radius (proximity) areas. Give the complete set of locations to target or exclude in one call; a location currently targeted but left out of the call is removed, and each location may be listed only once. Locations are identified by their numeric geo target constant id (e.g. 2840 for the United States); set a location's 'negative' flag to exclude it (for example target a country but exclude a state), and give targeted locations or proximities an optional bid_modifier between 0.1 and 10.0 (1.0 leaves bids unchanged; excluded locations may not have one). Re-running keeps the locations that stay in the set and updates their bid modifier, but a location cannot be switched between targeted and excluded in one call: omit it to remove it, then add it back the other way in a later call. A proximity area is a point (latitude, longitude) and a radius in miles or kilometers; proximity areas can only be targeted, not excluded, and are replaced as a whole on each call. Passing empty location and proximity lists clears the campaign's geo targeting so its ads can serve in all locations. To read the current geo targeting, query get_data with the criteria_type, campaign_criterion_location_geo_target_constant, campaign_criterion_negative, campaign_criterion_bid_modifier, campaign_criterion_proximity_geo_point_latitude_in_micro_degrees, campaign_criterion_proximity_geo_point_longitude_in_micro_degrees, campaign_criterion_proximity_radius and campaign_criterion_proximity_radius_units fields.
Set the geographic targeting of a Google Ads campaign: the countries, regions or cities its ads may serve in, plus optional radius (proximity) areas. Give the complete set of locations to target or exclude in one call; a location currently targeted but left out of the call is removed, and each location may be listed only once. Locations are identified by their numeric geo target constant id (e.g. 2840 for the United States); set a location's 'negative' flag to exclude it (for example target a country but exclude a state), and give targeted locations or proximities an optional bid_modifier between 0.1 and 10.0 (1.0 leaves bids unchanged; excluded locations may not have one). Re-running keeps the locations that stay in the set and updates their bid modifier, but a location cannot be switched between targeted and excluded in one call: omit it to remove it, then add it back the other way in a later call. A proximity area is a point (latitude, longitude) and a radius in miles or kilometers; proximity areas can only be targeted, not excluded, and are replaced as a whole on each call. Passing empty location and proximity lists clears the campaign's geo targeting so its ads can serve in all locations. To read the current geo targeting, query get_data with the criteria_type, campaign_criterion_location_geo_target_constant, campaign_criterion_negative, campaign_criterion_bid_modifier, campaign_criterion_proximity_geo_point_latitude_in_micro_degrees, campaign_criterion_proximity_geo_point_longitude_in_micro_degrees, campaign_criterion_proximity_radius and campaign_criterion_proximity_radius_units fields.
FAQs
What is the easiest way to connect Google Ads to ChatGPT?
Windsor MCP. Connect your Google Ads account at onboard.windsor.ai, open the Windsor app page in ChatGPT, and click Connect. Then activate it in any new conversation via + → More → windsor.ai. Your live Google Ads data is available in ChatGPT from that point on; no API credentials, no exports, no maintenance.
Does Windsor support Google Ads Manager Accounts (MCC)?
Yes. Windsor connects at the MCC level, so you can pull data from all sub-accounts into a single ChatGPT conversation. This is particularly useful for agencies that want to compare performance across client accounts, flag budget pacing issues, or roll up spend and ROAS across a managed portfolio.
Can ChatGPT see my search term report data?
Yes. Windsor pulls the search term report from the Google Ads API, giving ChatGPT access to the actual queries that triggered your ads, with spend, conversions, and conversion rate per query. This is one of the most valuable datasets for identifying negative keyword opportunities and search term overlap between match types.
How does Windsor handle Performance Max campaigns?
Performance Max is opaque by design, but Windsor pulls what the API makes available: asset group performance (cost, conversions, conversion value per asset group), product-level data where a Shopping feed is connected, and top-level campaign metrics. It’s not full transparency, but it’s enough to identify underperforming asset groups and flag products spending without converting.
Can I blend Google Ads data with my CRM or e-commerce store in ChatGPT?
Yes. Windsor connects 350+ data sources, and all of them are queryable in the same ChatGPT conversation. Common combinations: Google Ads + Shopify for revenue reconciliation, Google Ads + Salesforce or HubSpot for B2B pipeline attribution, and Google Ads + GA4 for post-click behaviour analysis. The join key is typically utm_campaign or utm_source passed through from the ad.
Can ChatGPT change campaigns in my Google Ads account?
With Windsor’s MCP, ChatGPT can manage your Google Ads account without switching to the Google Ads dashboard. You can: pause or enable ad groups and ads that are underperforming; set the bid strategy or adjust the max CPC on a campaign or ad group; push new negative keywords to cut wasted search terms; and create new campaigns, ad groups, and ads from scratch, with type, budget, and bid strategy set at creation. ChatGPT acts only on your direct instructions; nothing changes in your account unless you ask for it.
Does Windsor support Google Ads data for Shopping and YouTube campaigns?
Yes. Windsor pulls data across all Google Ads campaign types: Search, Shopping, Performance Max, Display, Video (YouTube), and Discovery. Campaign type is available as a dimension, so you can compare ROAS and CPA across types, or analyse Shopping product performance and YouTube view-through conversions independently.


