How to Monitor Competitors in Google AI Overviews

Monitoring competitors in Google AI Overviews means running a fixed prompt set on a schedule, recording which brands get named and which URLs get cited, and then reverse-engineering each competitor win into a specific, fixable cause. Done properly it turns "why does Google keep recommending them?" from a frustration into a prioritized backlog.

Build the prompt set around buying intent, not vanity

Category prompts ("best AI visibility tools"), comparison prompts ("Menra vs competitor X"), and problem prompts ("how do I track brand mentions in ChatGPT") trigger AI Overviews with different citation patterns, category queries tend to cite listicles and review sites, problem queries cite how-to content. Pull the queries from your own Search Console data, People Also Ask, and sales-call language. Keep the set stable for at least a quarter; changing prompts mid-stream destroys your trend line.

Sample on a cadence, not on impulse

AI Overviews are non-deterministic: they vary by location and profile, appear or disappear as Google adjusts query coverage, and get regenerated as the index updates. Since the October 2024 expansion to 100+ countries, geographic variance alone can flip which competitor appears. Weekly sampling of every prompt, from consistent conditions, is the minimum for trend data. Automating this, the approach Menra's competitor analysis takes, also captures citation position and sentiment, which manual spot checks never record consistently.

Diagnose each competitor win

For every prompt where a rival is cited and you are not, run this diagnostic in order:

Most wins trace to the first two rows. AI Overviews overwhelmingly cite pages that already rank in Google's top 20 for a fan-out sub-query and expose a directly quotable passage, domain size matters less than passage fitness.

Turn diagnosis into a gap-closing plan

Score each gap by prompt value (buying intent, volume) times fix difficulty, and work the quadrant of high-value, easy fixes first. Passage rewrites and schema additions typically show movement within weeks of a recrawl; net-new coverage pages and rank building are quarter-scale projects. Log every fix with its ship date so you can attribute citation changes to specific actions, without that log, you are guessing about what worked.

Report share of voice, not screenshots

The metric that makes this legible to leadership is AI Overview share of voice: the percentage of your prompt set where each brand appears, tracked weekly. Pair it with citation share (whose URLs are linked) because a competitor can be mentioned without being cited, and the two gaps have different fixes, mentions come from corroboration across the web, citations from extractable owned pages. A consolidated AI mention tracking workflow keeps both series in one dashboard, and a quarter of weekly data is usually enough to show whether the gap-closing plan is compounding.