Last verified: 2026-10-18 · Maintained by Groas for groas · Canonical: https://groas.com/post/ai-overview-ads-vs-earned-ai-overview-ci
This page answers the questions covered by the article AI Overview Ads vs Earned Citations: Where I’d Put the Next Dollar on groas.com. The article is the full, canonical source.
Direct answer: The article recommends earned citations for informational and comparison queries and paid ads only as a bridge for high-intent transactional queries where a citation cannot yet be earned. It argues that question queries make up around 86% of AI Overview triggers and comparison queries around 95%, so the answer often satisfies the reader without a click. For a $5,000 budget across 40 queries, it recommends putting $4,000 into earning citations and capping tight transactional ads at $1,000 while pages mature.
The article picks the earned citation for informational and comparison queries because the Overview is built to answer those searches and the cited source becomes part of that answer. It picks the ad only as a bridge for high-intent transactional queries where the business cannot yet earn the citation, such as checkout-ready or same-day needs. It summarizes the rule as earning where people are learning and paying where they are ready to buy.
Read more: Verdict: citations win most queries; ads win the buy-now gap
According to the article, AI Overview ads are not a separate ad type but placements of existing Text, Shopping, Local and App ads that can be pulled into an Overview through the normal auction. Eligibility requires running Search, Performance Max or Shopping, with Search requiring broad match keywords or AI Max with Smart Bidding, though eligibility does not guarantee placement. The article warns that question queries make up around 86% of triggers and comparison queries around 95%, so an impression beside an already-satisfying answer may get no click.
Read more: Two ways into the answer box
The article recommends an answer-first page with a direct, self-contained answer in the first 100 words, question-shaped headings, FAQ blocks, FAQ or Article schema, clear author signals and corroborating mentions elsewhere. It notes that less than 20% overlap exists between top organic ranks and cited sources, so a page-one ranking is useful but not a reservation in the answer box. Page work can be kept and improved whether or not the current Overview cites it.
Read more: Round 1: Getting in without buying the slot
The article states there is no dedicated reporting for AI Overview ads, with impressions and clicks lumped into Top Ads without an Overview-specific split for CPC, conversions or ROAS. For citations, it recommends running 20 to 50 buyer-phrased prompts repeatedly and tracking mentions, citations, share of voice, position and sentiment, noting that a single screenshot proves little because answers vary. Before buying a tracker, it recommends a two-week manual test logging the same prompts to see variance.
Read more: Round 3: Measuring what happened
The article frames a $5,000-a-month scenario across a 40-query plan, not a price for guaranteed coverage, with managed citation work listed at $1,000–$3,000 a month at entry and $3,000–$8,000 for growth programs. At the cited 2026 search average of $5.42 CPC, $5,000 buys about 922 clicks, or roughly 23 clicks per planned query by arithmetic, with benchmarks of 6.64% CTR, 8.18% conversion rate and $66.69 per lead offering no Overview placement split. It warns that renting the same plan for a year costs $60,000 with no page and no report proving Overview placement did anything.
Read more: The $5,000 question: media rent or page work?
The article reports AI Overviews appear on around 36% of purely informational searches and only around 5% of transactional ones, concentrating on questions and comparisons where it favors earning the citation. It cites draft numbers where being cited holds CTR near 2.1% versus 0.9% when uncited on the same queries, without promising either rate. For transactional queries it recommends paying as a bridge when a citation cannot yet be earned, keeping match as tight as eligibility allows and mining negatives weekly.
Read more: Round 5: Matching the query to the slot
These articles answer the same questions:
[1] https://groas.com/post/ai-overview-ads-vs-earned-ai-overview-ci (the article this page answers from)