Google Ads Reporting and Analytics: A Complete Guide

Google Ads carries the highest share of AI-assistant traffic of any advertising platform on Windsor.ai. That shows in what teams pull through it. Alongside the usual clicks, impressions and spend, conversions and the three impression-share fields sit right near the top of the list, ahead of quality score, keyword text, and most of what a classic PPC report leads with. Google Ads users query for decisions, not spend recaps.

The most common full pull is six fields together: campaign, clicks, conversions, date, impressions, spend. If you’re starting from scratch, that combination is the right first prompt.

Pull campaign, clicks, conversions, date, impressions, and spend for my
Google Ads account over the last 30 days, grouped by campaign. Sort by
spend, highest first.

This works the same whether you land it in ChatGPT, Claude, or Copilot once Google Ads is connected to Windsor.ai, or schedule the same pull into Looker Studio, Power BI, or BigQuery.

Campaign performance

campaign_status is one of the most requested fields on the connector, and it is usually there to filter, not just to display: teams narrow a report to ENABLED campaigns before they look at anything else. Add campaign_id alongside the canonical pull once you need to reference a specific campaign outside the report itself.

Pull campaign, campaign_status, clicks, conversions, date, impressions,
and spend for my Google Ads account over the last 30 days, only for
ENABLED campaigns. Sort by spend, highest first.

Conversions and value

conversions counts only the conversion actions marked to count toward your bidding, the ones your bid strategy optimizes for. all_conversions counts every conversion action regardless of that setting, so it is usually the larger number, and a wide gap between the two means conversions are happening that your bidding never sees. conversions_value and conversion_value return the same figure under two names, treat them as aliases rather than two different numbers. cost_per_conversion follows the same bidding-eligible definition as conversions.

Pull campaign, conversions, all_conversions, conversions_value, and
cost_per_conversion for the last 30 days. Where does all_conversions
run well ahead of conversions?

Search terms to negative keywords

search_term is one of the most-pulled fields on the connector, and it shows up in several of the most common multi-field combinations, alongside clicks and spend rather than on its own. On the write side, pushing negative keywords is the single most used write action on any Windsor connector: once you’ve reviewed a search-term pull for wasted spend, an assistant can turn the results straight into an exclusion list. Start with search_term, search_term_match_type, clicks, conversions, and spend.

Pull search_term, search_term_match_type, clicks, conversions, and
spend for the last 30 days. Which search terms have clicks but zero
conversions and should be added as negative keywords?

Impression share

Three fields, search_impression_share, search_budget_lost_impression_share, and search_rank_lost_impression_share, are requested by nearly as many teams as keyword text or quality score, more attention than the classic PPC glossary gives them. search_impression_share is the impressions you got divided by the impressions you were eligible for; the other two split the gap into what you lost to budget and what you lost to Ad Rank. search_absolute_top_impression_share narrows that to the single most prominent position. Reading budget-lost against rank-lost per campaign tells you whether the fix is more budget or a better ad and bid.

Pull campaign, search_impression_share, search_budget_lost_impression_share,
and search_rank_lost_impression_share for the last 30 days. Which
campaigns are losing more to budget than to rank?

Keywords and quality score

keyword_text and keyword_match_type are your live keywords, not Keyword Planner research data, a distinction the connector requires you to keep straight since Google Ads treats them as separate resources entirely. quality_score is a current snapshot, not history, and it sums across ad groups when ad_group is left out of the request, so a keyword scoring 3 in two ad groups comes back as 6. Always include ad_group when pulling it. See the Google Ads field-combination guide for this and the other cases where Google Ads rejects a field combination outright.

Pull campaign, ad_group, keyword_text, keyword_match_type, and
quality_score for the last 30 days, for ENABLED campaigns only.

Budgets and bidding

budget_amount and bidding_strategy_type are common pulls for anyone auditing spend allocation, alongside campaign_status to see what is eligible to spend. Setting a bidding strategy through an assistant is one of the least reliable write actions on this connector, worth a manual check before you trust that it ran. For a template that tracks budget pacing day by day, see the Google Sheets budget pacing template.

Pull campaign, budget_amount, bidding_strategy_type, and campaign_status
for my Google Ads account. Which enabled campaigns are closest to
their daily budget cap?

Account and campaign health

campaign_primary_status_reasons explains why a campaign is not serving, or not serving optimally, and ad_group_ad_policy_summary_approval_status does the same for individual ads Google has flagged. Both are worth pulling before spending time debugging a campaign that Google Ads has already explained.

Pull campaign, campaign_status, campaign_primary_status_reasons, and
ad_group_ad_policy_summary_approval_status for the last 7 days. Flag
anything not serving normally.

Cross-channel: Google Ads, Meta Ads, and GA4

A large share of Google Ads accounts also pull Meta Ads in the same week, more than any other cross-connector pairing on this platform, with GA4 close behind and Search Console a smaller but consistent third. Comparing Google Ads and Meta Ads directly, on one blended report, is common enough to be worth doing properly: see Facebook Ads vs Google Ads: One Report for Both Platforms. For a GA4-joined view of what a lead or purchase costs end to end, see Google Ads CPA Tracking in BigQuery (with GA4).

Campaign management from an assistant

Pushing negative keywords is the single most used write action on any Windsor connector, and it runs reliably. Pushing new keywords, creating ad groups, and creating campaigns all run about as well. Setting a bidding strategy is the exception: it fails more often than any other Google Ads write action, so confirm the result rather than assume it took. See How to Connect Google Ads to Claude for the full set of write actions, or How to Manage Google Ads Campaigns from ChatGPT for the ChatGPT version.

Google Ads API without code

Everything above uses Google Ads’ real reporting fields without writing any code. Connect your account to Windsor.ai once, and ChatGPT, Claude, or Copilot can query it directly through plain-language prompts, across all 2,542 fields the connector exposes.

Fields that can’t be combined

One error family accounts for more failed Google Ads pulls than any other on this connector: fields Google Ads simply will not return together in one request. Keyword research fields mixed with live performance, ad-group-criterion keyword fields mixed with standard metrics, conversion-action metadata mixed with campaign-level stats, and auction-insight segments mixed with anything else are the recurring cases. The fix is always the same: split the pull into two requests and join the results yourself, by date and campaign. See the full breakdown, with the fix for each case, in the Google Ads “Fields Cannot Be Queried Together” Error: Troubleshooting Guide.