Reporting for AI Max for Search campaigns

AI Max for Search campaigns offers several reporting views, each with a different level of granularity and focus.

Dedicated view: ai_max_search_term_ad_combination_view

This is the most granular report available for AI Max for Search campaigns. It shows the exact combination of user intent (the search term) and dynamic assets (headline and landing page) that the system used for the expansion.

  • Key fields:
    • search_term: The query the user entered.
    • landing_page: The specific (often dynamically selected) URL the user was sent to.
    • headline: The concatenated headline assets used in the ad.
  • Usage: Best for auditing the relevance of AI-driven matches and seeing which "ad recipes" are converting.

Search term attribution: search_term_view

To isolate the impact of AI Max, use the standard search term report and filter on search_term_match_source.

  • AI_MAX_KEYWORDLESS: Matches derived from your website content or landing pages, requiring no keyword.
  • AI_MAX_BROAD_MATCH: Matches that expanded upon your existing keywords using AI-driven broad match logic.

Landing page view: expanded_landing_page_view

Use this view to identify URLs favored by AI Max expansion.

  • Filter: campaign.ai_max_setting.enable_ai_max = TRUE
  • Field: expanded_landing_page_view.expanded_final_url

How to interpret the reports

When analyzing these reports, focus on the following three signals:

  • Incremental volume (expansion efficiency): Compare metrics for AI_MAX_KEYWORDLESS with ADVERTISER_PROVIDED_KEYWORD. If the conversion rate (CVR) for Keywordless is comparable to your keywords but at a lower Cost Per Acquisition (CPA), the AI is successfully finding "cheap" incremental reach that your manual keyword list missed.
  • Match intent versus asset relevance: In the ai_max_search_term_ad_combination_view, examine the headline items. Because these headlines are often concatenated or selected by AI, a mismatch between the search_term and the headline indicates where you may need to provide more diverse or relevant assets to the campaign.
  • Landing page performance: If certain landing_page URLs in the report have high bounce rates or low conversions, it may indicate that the AI is over-indexing on specific pages of your site that are not optimized for conversions. You can use this data to refine your site or provide exclusion rules.

Example GAQL

This query returns search terms specifically driven by AI Max expansion.

SELECT
  campaign.id,
  campaign.name,
  search_term_view.search_term,
  segments.search_term_match_source,
  metrics.impressions,
  metrics.clicks,
  metrics.conversions
FROM search_term_view
WHERE
  segments.search_term_match_source IN ('AI_MAX_KEYWORDLESS', 'AI_MAX_BROAD_MATCH')
  AND segments.date DURING LAST_30_DAYS

How to interpret the campaign numbers

  • Why the combined sum (AI Max + Broad) exceeds campaign clicks: This is due to double counting.

    • The overlap: The ai_max_search_term_ad_combination_view and search_term_view are not mutually exclusive. Many search terms generated by AI targeting are reported in both views.

    • Recommendation: Never sum metrics from these two views. They represent different reporting dimensions for the same underlying traffic.

  • Why the combined sum (AI Max + Broad) falls short of campaign clicks: This is due to privacy thresholds and filtered match types.

    • Privacy thresholds: Some search terms that do not have enough query activity are omitted from the search terms report to maintain our standards on data privacy. Clicks from these "hidden" terms are included in the campaign total but are excluded from the search term reports (search_term_view and ai_max_search_term_ad_combination_view).

How to correctly report

If you need a breakdown of performance by how traffic was targeted, use a single query on the search_term_view and group by segments.search_term_match_source.

Don't use the ai_max_search_term_ad_combination_view for reconciliation, as it is a specialized view for ad-combination testing, not a financial reconciliation tool.

AI Max exclusion rules

With AI Max, exclusion rules restrict Keyword Expansion (Broad Match/Keywordless) or landing page selection.

Exclude search terms (negative keywords)

  • Target: Irrelevant search terms identified in the ai_max_search_term_ad_combination_view.
    • Level: Ad Group or Campaign.
  • Resource: CampaignCriterion / AdGroupCriterion
  • Fields: keyword.text (Negative Exact/Phrase/Broad), keyword.match_type, and negative = TRUE.

Exclude landing pages (URL exclusions)

  • Target: Low-converting or irrelevant URLs, for example, /blog/, /login/, or /thank-you/.
  • Level: Campaign.
  • Resource: CampaignCriterion
  • Fields: webpage.conditions (URL_CONTAINS, PAGE_TITLE_CONTAINS) and negative = TRUE.

Campaign-level brand exclusions

  • Target: Competitor or own brand names.
  • Level: Campaign.
  • Resource: CampaignCriterion
  • Fields: brand_list (referencing a BrandList resource) and negative = TRUE.

Note: Use ai_max_search_term_ad_combination_view to audit performance and identify specific landing_page URLs or search_term strings that require exclusions.

When to use exclusions

Use the following guidance to decide which type of exclusion is best for your situation.

Negative keywords (search term exclusions)

When to use: When the user's intent (the query) is fundamentally irrelevant or unprofitable, regardless of which page they land on.

  • Scenario A (poor intent): Your ai_max_search_term_ad_combination_view shows matches for "free", "cheap", "DIY", or "how to" when you sell premium enterprise software.
  • Scenario B (irrelevant expansion): The AI_MAX_BROAD_MATCH expansion goes too far. For example, you sell "water filters" and the AI expands to queries about "pool cleaning services".
  • Strategy: Apply as Exact match if the specific query is problematic, or Phrase/Broad if an entire category of queries (that is, anything containing "free") needs to be blocked.

URL exclusions (negative webpage targets)

When to use URL exclusions: When the AI successfully finds relevant search terms but routes users to the wrong page on your website, or when a specific page consistently yields a high CPA or low conversion rate.

  • Scenario A (non-commercial pages): AI Max keywordless expansion sends paid traffic to your /careers, /privacy-policy, or /customer-support-login pages. Exclude these URLs immediately.
  • Scenario B (the "whack-a-mole" problem): You notice in the ai_max_search_term_ad_combination_view that dozens of different, seemingly acceptable search terms all result in bounces because they land on an outdated blog post. Instead of adding dozens of negative keywords, add one URL exclusion for that blog post.
  • Strategy: Use URL_CONTAINS for directories (for example, URL_CONTAINS 'blog') or exact URL exclusions for specific problematic pages.

Campaign-level brand exclusions

When to use campaign-level brand exclusions: When strict traffic segmentation is needed to prevent AI Max from cannibalizing your existing branded Search campaigns.

  • Scenario A (protecting brand CPCs): You already have a highly optimized Exact Match campaign for your own brand name. If you enable AI Max on a generic non-brand campaign to find incremental reach, and you notice AI Max is starting to bid on your brand name, apply a brand exclusion list containing your own brand to force AI Max to only find net-new, generic queries.
  • Scenario B (competitor policy): Your company has legal agreements not to bid on specific competitor brand names, and a fail-safe is needed to ensure that AI Max's broad expansion never triggers ads on those terms.

Summary guidelines

  • Bad query = negative keyword.
  • Bad destination = URL exclusion.
  • Bad cannibalization = brand exclusion.