AI search opportunities

Bloomiro now turns your saved Google AI Overview checks into a focused Opportunities workspace.

It uses the evidence already captured for your prompts. It does not start a new search when you open the page, and reading the results does not use credits.

Open opportunities

Open opportunities shows the latest measured opportunity for each prompt. This keeps the list useful when you have checked the same prompt many times.

Each opportunity explains:

  • The exact search used
  • The page Bloomiro found in your organic results
  • Whether that page appeared anywhere in Google’s captured top 100 organic results for that exact search prompt
  • Whether your site was cited in the AI Overview
  • Which competitors or other sources were cited
  • The related questions and searches connected to the same intent
  • Why the item was surfaced and what to investigate next

The main opportunity types are:

  • Near win: your page ranks in the captured top 20 but is not cited in the AI Overview.
  • Competitor gap: a tracked competitor is cited while your site is not.
  • Off-SERP citation: Google cites a page that was not found in the captured top 100 organic results for that search prompt.
  • Question cluster: People Also Ask and related searches reveal a connected group of questions that may belong on one strong page.

These are evidence-backed hypotheses. They are not guarantees of a ranking increase or an AI citation.

Repeated patterns

Repeated patterns aggregates evidence across saved checks instead of showing one prompt at a time.

Search ideas

Search ideas are repeated wording from People Also Ask and related searches.

People Also Ask is the question box Google displays for some searches. Related searches are alternative searches Google displays around the original search. Both are useful signals about nearby questions and intent, but neither proves search volume.

Use them to improve the relevant page, expand an existing section, or decide whether a genuinely different intent deserves its own page. Do not create thin pages for every phrase.

Cited winners

Cited winners shows pages and domains that Google repeatedly used as AI Overview sources across your saved checks.

This helps you study which sources Google selected and whether the pattern is concentrated in a domain, a type of source, or a particular prompt. A repeated citation is observed evidence, not proof that the page is authoritative or that copying it will earn a citation.

Prompt history

Prompt history lets you inspect the saved checks behind the current opportunity view. It helps you see when a prompt was last checked and keep the latest result separate from older observations.

The workspace stays bounded. It does not load every historical AI answer into the page or into an AI assistant context.

Content patterns

Content patterns is an optional deeper analysis of cited pages. Bloomiro selects the top 20 exact pages that were cited in at least two saved Google AI Overview checks. If fewer pages qualify, it analyzes only those pages.

It measures recurring topics and phrases separately in:

  • Page titles
  • Meta descriptions
  • Headings
  • Main body content

Each result shows how many cited pages contain the phrase, how many domains are represented, and where the phrase appears. You can open one pattern to see the exact cited source pages and locations.

The analysis runs in the background, so you can leave the page and return later. A new analysis replaces the previous Content patterns result with the newest eligible cited pages.

Actions and MCP

Actions and MCP use the same project-scoped evidence.

The assistant starts with a compact overview, then reads a bounded list, one opportunity, one repeated pattern page, or one content pattern only when the question requires it. It does not receive raw AI answer dumps, full SERP exports, or every historical run by default.

Use get_ai_search_opportunities with:

  1. view=overview for counts and the available next reads.
  2. view=list for a short ranked list of the latest opportunity per prompt.
  3. view=detail with an opportunity_id for one exact opportunity.
  4. view=patterns with pattern_type=search_idea or cited_source for repeated evidence.
  5. view=content_patterns with a content_pattern_id for one pattern and its cited pages.

All read views are free. Starting the cited page analysis is separate background work and costs one credit when it completes successfully.

Credit cost

The Billing page now lists Content patterns scan: 1 credit per scan. Opening Opportunities, reading repeated patterns, viewing prompt history, and using the read-only MCP or Actions views do not use credits.