How Does Google Ads Work? The Auction, Plainly

Last updated: September 13, 2026

1. What Actually Happens When Someone Searches

The moment a person types a query into Google and presses Enter, a process that takes roughly 200 milliseconds — less time than an eye-blink — runs between their keystroke and the ads they see. Understanding that process is what this article is about.

Here is the sequence, in plain English:

1. The person’s browser sends the query to Google’s servers.
2. Google identifies every advertiser whose keywords could match that query.
3. For each eligible advertiser, Google runs a calculation called an Ad Rank — a number that reflects both how much the advertiser is willing to pay and how useful Google predicts their ad will be to this person, in this moment, on this device, in this location.
4. Google sorts the eligible advertisers by their Ad Rank scores and assigns positions: highest score gets first position, next gets second, and so on — except that an advertiser must clear a minimum threshold score to appear at all, and a higher threshold to appear above the organic results.
5. Google calculates what each advertiser actually pays if their ad is clicked. This is not their bid. It is the minimum they needed to pay to beat the advertiser directly below them in the ranking.
6. The page renders, the ads appear, and the user sees results — all in those 200 milliseconds.

Two facts from that sequence do the most work in this article. First, the advertiser with the highest bid does not automatically get the top position. Quality matters at least as much as bid, and sometimes more. Second, every advertiser in every auction pays less than their maximum bid — often considerably less — because the price is set by the competition, not by the ceiling they set themselves.

Everything else in this article — Ad Rank, Quality Score, actual CPC, Smart Bidding, and what you can actually change — is an expansion of those two facts. The same four advertisers are used throughout as a running example so the numbers stay consistent from section to section.

2. What Google Ads Is, in One Section

Google Ads (formerly Google AdWords) is Google’s paid advertising platform. Advertisers pay to show ads to people using Google Search, Google’s partner websites, YouTube, Gmail, Google Maps, and the Google Display Network — a collection of millions of third-party websites and apps that have agreed to show Google ads.

The platform matters to this article because the auction mechanics described here apply fully only to Search campaigns — ads that appear when someone types a query into Google. Other surfaces have their own relevance signals and ranking systems, which are noted in the table below. Quality Score as a 1–10 diagnostic metric is a Search concept; Display, YouTube, and Shopping use analogous quality signals, but they are not called Quality Score and they do not surface the same way in reporting.^[1]

The intent of the person seeing the ad varies sharply by surface, and that intent difference is the reason advertisers choose one surface over another. Search captures explicit demand — someone is already looking for something. Display and YouTube create or accelerate demand among people who have not searched yet.^[1] That distinction changes what a “good” ad looks like, what a realistic conversion rate is, and how budget should be allocated.

For the rest of this article, “the auction” means the Search auction unless stated otherwise. Performance Max, AI Overviews, and Shopping each deserve their own treatment; the Omologist library covers them separately.

3. The Auction, Step by Step

The Google Search auction runs independently for every single query. The competitors, the bids, the quality scores, and the price each advertiser pays can all be different from one search to the next, even for the same keyword searched by the same person an hour apart.^[3] What follows is the step-by-step walk-through of one auction instance.

Step 1: Eligibility check

Before any ranking happens, Google filters out advertisers whose ads are disapproved, whose targeting does not match the query context, or whose Ad Rank falls below the minimum threshold required to show on the page at all.^[1][4] An Ad Rank threshold — the minimum Ad Rank score an ad must reach to be eligible for a given position — varies by query, device, location, and time of search. It is not a single fixed number; Google recalculates it for each position in each auction.^[5] Advertisers below the threshold for the page do not enter the auction for that query.

Step 2: Ad Rank calculation

Ad Rank is the number Google uses to determine whether an ad is eligible to show and, if eligible, where it appears relative to other ads.^[1] The full calculation uses six inputs described in section 4. For teaching purposes, a simplified version of the formula is:

Ad Rank (simplified) = Max CPC bid × Quality Score

This is labelled simplified because the real auction also incorporates Ad Rank thresholds, the expected impact of ad assets, auction context signals (device, location, time, query intent), and competitiveness — none of which appear in the multiplication above.^[1][3]

Step 3: Position assignment

Google sorts eligible advertisers from highest to lowest Ad Rank. The highest Ad Rank earns position 1 (the top ad slot above the organic results), the second highest earns position 2, and so on. An advertiser with a high Ad Rank can appear at the top even with a lower maximum bid than a competitor who scores lower on quality.^[3]

Step 4: Actual CPC calculation

The price each advertiser pays is not their maximum bid. It is the minimum amount needed to beat the Ad Rank of the advertiser directly below them.^[1] The simplified formula used throughout this article is:

Actual CPC (simplified) = (Ad Rank of the advertiser below ÷ your Quality Score) + $0.01

This is simplified: in reality, reserve prices, thresholds, and format adjustments can alter the exact figure.^[1][3] But the formula is accurate enough to understand why a better Quality Score lowers what you pay — it is in the denominator.

The four advertisers used throughout this article

The same four advertisers appear in every section of this article so the numbers stay consistent. Their values are fixed — bids, scores, and positions do not change unless a section explicitly states that one figure has changed and by how much, and then returns to these values.

Notice that Advertiser B bids the most ($6.00) but places third, behind Advertiser A ($4.00 bid) and Advertiser C ($3.00 bid). The Quality Score difference is the reason. This point is expanded in section 8.

4. Ad Rank: The Six Inputs

Google’s official documentation states that Ad Rank is calculated from six factors.^[1] This section translates each one into plain language and shows what, if anything, you can do about it. The table below summarizes all six; the prose after it explains the ones that carry the most practical weight.

Bid: necessary but not sufficient

Bid is the most visible lever and the easiest to raise, which is why beginners often treat it as the primary tool. It is not. Bid raises Ad Rank linearly in the simplified model, but quality can move Ad Rank by larger amounts. Advertiser B in the four-advertiser table bids 50% more than Advertiser A ($6.00 versus $4.00) and still places third because B’s Quality Score (4) is half of A’s (8). Raising B’s bid further would be expensive; raising B’s Quality Score from 4 to 8 would double Ad Rank without spending another dollar on CPC.

Auction-time ad quality

This is the input that produces the Quality Score diagnostic number you see in the interface, but the two are not the same thing. Auction-time ad quality is evaluated in real time for each search, using the three components — expected CTR, ad relevance, and landing page experience — as signals.^[3] The Quality Score (a 1–10 number you see in the Keywords column) is a periodically updated reporting metric that estimates how your quality compares with competitors; Google explicitly says it is not the auction input itself.^[6] Sections 5 through 7 cover each component. Section 10 covers the diagnostic number.

Ad Rank thresholds

The threshold concept surprises many beginners: you can have a positive bid and an approved ad and still not appear for a query because your Ad Rank does not clear the minimum required for that query’s context.^[5] The threshold for appearing in top positions (above the organic results) is generally higher than the threshold for appearing at the bottom of the page.^[2] This is one reason top-position CPCs are higher — not only are more advertisers competing for those slots, but Google’s threshold for entering them filters out lower-quality participants.

Context of the search

Google recalculates Ad Rank for each auction instance, which means the same keyword can produce a different Ad Rank depending on whether the search came from a mobile phone in Chicago at 8 p.m. or a desktop in rural Nebraska at 2 a.m.^[1][3] The competitor set, the thresholds, and the quality signals all vary by context. This is also why average CPC figures by industry are blunt instruments — your actual CPCs vary within a single campaign by time of day, device, and location.

Expected impact of assets and formats

Assets — previously called ad extensions — include sitelinks (extra links below the ad), callout text (short phrases highlighting offers), structured snippets (lists of features), call extensions, image extensions, and others. Google predicts whether showing an asset for this particular query will improve the likelihood of a click.^[1] When that prediction is positive, the asset raises Ad Rank. This means two advertisers with identical bids and Quality Scores can have different Ad Ranks if one has richer, more contextually relevant assets.

5. Quality Score Component 1: Expected CTR

Expected click-through rate (Expected CTR) — the likelihood that your ad will be clicked when shown for a given keyword, measured relative to other advertisers competing on that same term — is the first of three components that make up your Quality Score.^[6] Google evaluates it at auction time, using historical performance data for your keyword and ad combination alongside signals from other advertisers, and then reports a simplified version back to you as a keyword-level status: Above average, Average, or Below average.^[6]

The important word in that definition is expected. Google is not looking only at how many clicks your ad has received so far. It is predicting how likely it is to get clicked, based on the message, the keyword, the competitive context, and accumulated data about how similar ads perform in similar auctions.^[3] This means a brand-new ad on a keyword can receive an Expected CTR status before it has generated a single impression of its own, because Google draws on broader auction data.

Expected CTR is also not the same thing as Ad Strength. Ad Strength is a creative-completeness indicator for Responsive Search Ads — it tells you whether your ad has enough headline and description variety — but it carries no demonstrated correlation with actual CTR or conversion rate, and it is not an Ad Rank input.^[7][8] For a detailed breakdown of Ad Strength, see the Omologist article on Ad Strength.

What moves Expected CTR

Three things have the most leverage over your Expected CTR status.^[9][6]

• Message match between keyword and ad headline. When a searcher’s query appears — or closely mirrors — the first headline of your ad, Google’s prediction of a click goes up. Searchers are pattern-matching what they typed against what they see on the results page. If your headline answers the query directly, the prediction improves.
• Offer specificity and relevance. Generic headlines like “Learn More” or “Quality Service” suppress predicted CTR. Specific offers — a price, a timeframe, a benefit the competitor’s ad does not state — raise it.
• Historical ad performance on that keyword. If your ad has a strong CTR track record for a keyword across past auctions, that history feeds forward into future Expected CTR estimates. This is one reason a well-established campaign with real data is harder for a new entrant to displace purely on bid.

What does not move Expected CTR on its own: raising your bid, increasing your budget, or improving your landing page. Those factors affect other parts of Ad Rank and Quality Score, but Expected CTR is purely a prediction about the ad and the keyword — not the page behind the ad and not the amount you are willing to spend.^[6]

When to pull it, and when not to

What it changes: Improving Expected CTR raises your Quality Score, which feeds into the Ad Rank calculation and can lower your actual CPC for the same position.

Why it works: Because actual CPC is set partly by the Ad Rank of the advertiser below you divided by your own Quality Score (simplified formula), a higher Quality Score shrinks the amount you pay per click even if your bid and your competitor’s Ad Rank stay the same.^[3]

When to use it: When your Expected CTR status column shows Below average for any keyword driving meaningful spend — specifically any keyword with more than 200 impressions in the last 30 days — rewriting the associated headlines is the highest-leverage single action available.

When not to: If your Expected CTR shows Average or Above average and your Ad Relevance or Landing Page Experience shows Below average, chasing Expected CTR further is the wrong priority. Fix the lower-scored component first; that is where the Quality Score ceiling is being set. Rewriting headlines repeatedly when the problem is landing page mismatch wastes time and can destabilize copy that is already performing.

6. Quality Score Component 2: Ad Relevance

Ad relevance measures how closely your ad’s message matches the intent behind the user’s search query and the keyword triggering the ad.^[6] Where Expected CTR asks “will this ad get clicked?”, Ad Relevance asks “does this ad address what the person actually searched for?” The two are related but distinct: an ad can have a compelling, click-worthy headline that is still too generic to score well on relevance, because relevance is about topical fit, not just persuasiveness.

Google reports Ad Relevance as one of the three component statuses — Above average, Average, or Below average — at the keyword level.^[6] A Below average status is a direct signal that the connection between your keyword, your ad text, and the searcher’s intent is too loose.

Why tight ad groups exist

The primary structural reason to build tightly themed ad groups — grouping keywords by a single, focused topic rather than mixing related-but-different terms — is Ad Relevance. When one ad must serve twenty different keyword intents, no single headline can match all of them precisely. The result is Below average Ad Relevance for most of the keywords in that group, which drags Quality Score down across the board.

The practical standard most practitioners use is to keep ad groups narrow enough that every keyword in the group could plausibly appear, verbatim or very close to it, in at least one ad headline. If a keyword requires a meaningfully different headline than the others in the group, it belongs in a different ad group with its own dedicated ad.^[9]

This is not just a tidiness principle. Because Ad Relevance affects Quality Score, and Quality Score affects actual CPC, a bloated ad group with mixed intent costs more per click on every keyword it contains — not because the bids are wrong, but because the quality signal is diluted.

What moves Ad Relevance

• Splitting mixed-intent ad groups. Separating “emergency plumber” and “plumber cost estimate” into two ad groups, each with its own dedicated ad copy, immediately raises relevance for both.
• Mirroring keyword language in ad text. Ad text does not need to be a verbatim repetition of the keyword, but the language family and intent should match. A keyword about “24-hour plumbing repair” and an ad headlined “Affordable Plumbing Services” are misaligned on urgency, even though both are about plumbing.
• Removing keywords that do not fit the ad group’s theme. A keyword that cannot be served by any headline in the ad group without sounding forced is dragging Ad Relevance down. Move it or remove it.

When to pull it, and when not to

What it changes: Improving Ad Relevance raises Quality Score and can increase auction-time ad quality, making the same bid more competitive and reducing actual CPC.

Why it works: Google’s auction-time quality evaluation includes Ad Relevance as an explicit component.^[3][6] When a keyword and its ad text address the same intent, Google’s model predicts a better user experience, which translates into a higher quality signal in the auction.

When to use it: When the Ad Relevance column shows Below average on any keyword spending more than $50 in the past 30 days, or when your ad group contains more than 10 to 15 keywords with meaningfully different intent. Open the Keywords view, add the Ad Relevance column (see section 10 for the column-adding procedure), and sort by status.

When not to: If all keywords in a group already show Average or Above average Ad Relevance, restructuring ad groups is not the constraint. The constraint is elsewhere — check Expected CTR and Landing Page Experience statuses first. Splitting ad groups that already score well on relevance adds account complexity without improving the Quality Score ceiling.

7. Quality Score Component 3: Landing Page Experience

Landing page experience measures how relevant, useful, and trustworthy Google considers the page a user reaches after clicking your ad.^[6] It is the third Quality Score component and the one most frequently overlooked, because it exists outside the Google Ads interface — on your website — where most advertisers feel less immediate control.

Google evaluates landing page experience at auction time using a combination of signals that estimate whether the page is likely to satisfy the searcher who arrives on it.^[3][6] The key signals Google’s documentation points to are: how closely the page content relates to the ad and keyword, how transparent the page is about what the business offers, how easy the page is to navigate, and how well it loads and functions on the device the searcher is using.^[10]

For a deeper treatment of landing page optimization — page speed, content structure, trust signals, and conversion rate — see the Omologist article on landing pages.

What Google actually measures

Google’s public guidance identifies four main dimensions of landing page experience.^[10]

• Relevant and original content. The page should directly address the search intent that triggered the ad. If an ad promotes “emergency roof repair in Austin,” the landing page should be about emergency roof repair in Austin — not the company’s general services homepage.
• Transparency and trustworthiness. The page should clearly identify the business, what it offers, and how to contact it. Hidden fees, unclear ownership, or a lack of basic trust signals (phone number, address, privacy policy) are negative signals.
• Ease of navigation. Google’s evaluation considers whether the page makes it straightforward to find the information the user came for, without requiring the user to hunt through unrelated content.
• Mobile usability and load speed. Pages that load slowly or break on mobile devices receive lower landing page experience scores, because Google factors in device context at auction time.^[3]

What Google does not publicly confirm is the exact weighting of each signal, or the precise technical thresholds for page speed that trigger a score change. The documented principle is that the page must be useful for the specific search that triggered the ad — not useful in general, but useful for that query.

The message-match problem

The single most common cause of Below average Landing Page Experience is a mismatch between what the ad promises and what the landing page delivers. An ad that says “50% off HVAC tune-ups this October” but lands on a homepage that says nothing about a promotion will score poorly on relevance, because Google predicts the visitor will not find what they came for and may leave quickly. Directing every ad to the homepage instead of a dedicated landing page is the structural version of this error.

When to pull it, and when not to

What it changes: Improving Landing Page Experience raises Quality Score, which can lower actual CPC and improve Ad Rank without touching your bid.

Why it works: Landing Page Experience is an explicit Quality Score component that feeds into auction-time ad quality.^[3][6] A higher quality signal means your ad competes more effectively against higher-bidding advertisers with weaker pages.

When to use it: When the Landing Page Experience column shows Below average for any keyword with more than 100 clicks in the past 30 days, or when your site sends all ad traffic to the homepage regardless of keyword. Both are clear signals that landing page alignment is the binding constraint on Quality Score.

When not to: If Landing Page Experience already shows Average or Above average across your keywords, further landing page work will not meaningfully move Quality Score. The constraint has shifted to Expected CTR or Ad Relevance. Additionally, if your campaign is running on Smart Bidding with strong conversion data, Google’s auction-time bidding already partially compensates for quality differences at the bid level — landing page fixes still matter for conversion rate, but they may not be the fastest path to lower CPCs in a mature Smart Bidding campaign.

8. Why the Highest Bidder Loses

Section 3 introduced the four advertisers and their Ad Rank positions. This section explains the mechanism behind those positions — specifically, why Advertiser A wins the top spot with a $4.00 bid when Advertiser B bids $6.00 and still lands in third place. The answer is in the relationship between bid, Quality Score, and actual CPC.

The Ad Rank table, revisited

The Ad Rank formula used here is simplified: in this illustration, Ad Rank = Bid × Quality Score. The real auction adds thresholds, format impact, context signals, and competitiveness adjustments that the simplified formula does not capture.^[1][3]

The actual CPC formula is also simplified: Actual CPC = (Ad Rank of the advertiser directly below you ÷ your Quality Score) + $0.01. Google describes the mechanism as “pay just enough to beat the advertiser below you,” subject to reserve price rules and thresholds.^[11] The formula is a practitioner approximation, not an officially published equation.

Walking through the arithmetic

Using the simplified formula for each advertiser:^[11]

• Advertiser A (1st): Ad Rank below = C’s 27. Actual CPC = (27 ÷ 8) + $0.01 = $3.375 + $0.01 = $3.39. A bids $4.00 but pays $3.39.
• Advertiser C (2nd): Ad Rank below = B’s 24. Actual CPC = (24 ÷ 9) + $0.01 = $2.667 + $0.01 = $2.68. C bids $3.00 but pays $2.68.
• Advertiser B (3rd): Ad Rank below = D’s 12. Actual CPC = (12 ÷ 4) + $0.01 = $3.00 + $0.01 = $3.01. B bids $6.00 but pays $3.01.
• Advertiser D (4th): No advertiser below D, so D pays the reserve price — the auction floor Google sets for the query. The reserve price is not publicly disclosed and varies by context.^[1][5]

Notice what just happened. Advertiser B spent $6.00 — 50% more than A’s $4.00 — and landed in third place, paying $3.01 per click. Advertiser C spent $3.00 and landed in second place, paying only $2.68 per click. Advertiser C is paying less per click than B and sitting one position higher, entirely because C’s Quality Score of 9 versus B’s 4 more than compensates for the bid gap.

This is the core insight the auction is designed to produce: Google rewards advertisers who write relevant ads and send traffic to useful pages, not simply those who bid the most. The reward is a lower actual CPC for the same or better position.

What this means in practice

A $2.00 increase in Quality Score — for example, moving from 6 to 8 — reduces your actual CPC by the same ratio as increasing the denominator in the simplified formula. For Advertiser B, if Quality Score improved from 4 to 8 (while all other values stayed fixed), Ad Rank would jump from 24 to 48, landing B in first position ahead of A, and the actual CPC calculation would change entirely. Quality Score improvement is not just a cosmetic metric — it directly reprices every click.

The honest caveat: the real auction is more complex than the simplified formula suggests. Ad Rank thresholds mean an ad must clear a minimum bar to appear at all, and a separate, higher threshold applies to top-of-page positions.^[1][4][5] Context — device, location, time of day, query — shifts both thresholds and competitor sets in every individual auction.^[3] The simplified formula gives you the right intuition; the real auction adds friction that the formula smooths away.

9. What You Actually Pay

Your max CPC (maximum cost-per-click) — the highest dollar amount you are willing to pay for a single click — is your bid ceiling, not your bill. Google charges you the actual CPC, which is the amount you are actually charged when a click happens, and it is almost always lower than your max CPC.^[11] The difference is not rounding: it is the auction’s built-in efficiency mechanism, the same one that makes Quality Score financially meaningful.

Google’s official explanation is that you pay the minimum required to hold your position — specifically, the amount needed to beat the Ad Rank of the advertiser immediately below you, subject to Ad Rank thresholds and any reserve price floor.^[3] The simplified practitioner formula is:

Actual CPC (simplified) = (Ad Rank of advertiser below you ÷ your Quality Score) + $0.01

Label that simplified: the real auction adds thresholds, format-impact adjustments, reserve prices, and context signals that can shift the charge above the naive formula result.^[1][3] Still, the simplified version is accurate enough to make the key point: a higher Quality Score divides the denominator harder, so you pay less for the same position.

The four-advertiser scenario introduced in section 3 makes this concrete. The fixed values are: Advertiser A bids $4.00, Quality Score 8, Ad Rank 32; Advertiser B bids $6.00, Quality Score 4, Ad Rank 24; Advertiser C bids $3.00, Quality Score 9, Ad Rank 27; Advertiser D bids $2.00, Quality Score 6, Ad Rank 12. Positions run A first, C second, B third, D fourth.

Now suppose Advertiser D raises its bid from $2.00 to $5.00. Its Quality Score stays at 6, so its new simplified Ad Rank is 5.00 × 6 = 30. That pushes D above B (Ad Rank 24) and above C (Ad Rank 27), landing D in second place. Positions become: A first, D second, C third, B fourth. The bids, scores, and Ad Ranks of A, B, and C have not changed — only D’s bid changed.

After D’s bid raise, D moves up but pays $4.51 — more than A pays at position 1 ($3.76), because D’s Quality Score of 6 divides less efficiently than A’s score of 8. C’s actual CPC is unchanged at $2.68, because the advertiser below C (now B at Ad Rank 24) has not moved. All values for A, B, and C return to their original figures for every other section of this article.

CPC also varies by the specifics of each individual auction, independent of competitors. The same keyword can produce a different charge depending on the device the searcher uses, the hour of the search, the searcher’s location, and the exact query string — because each of those factors changes which competitors enter the auction, what thresholds apply, and what Quality Score Google estimates for your ad in that moment.^[1][3] A desktop search at 9 a.m. in Chicago and a mobile search at 11 p.m. in rural Illinois are two separate auctions, even for the same keyword.

10. The 1-10 Quality Score Column

The Quality Score column in Google Ads is a 1-to-10 diagnostic metric reported at the keyword level. Google is explicit that it is not the signal the auction uses in real time; it is a periodic snapshot that estimates how your ad quality compares with other advertisers competing on the same keyword.^[6] The number you see this morning may not reflect a landing-page change you deployed yesterday, because the reported score updates on its own schedule rather than instantly.^[6]

Each overall score is built from three component statuses, each reported as Above average, Average, or Below average relative to other advertisers who showed ads for that keyword:^[6]

• Expected CTR — the likelihood that your ad is clicked when shown for this keyword, compared with competitors.
• Ad Relevance — how closely your ad message matches the intent of the keyword.
• Landing Page Experience — how relevant and useful your landing page is to someone who clicks.

A “Below average” flag on any component is the most actionable signal the column offers: it tells you which of the three levers is pulling your effective auction-time quality down, even if the overall score looks acceptable. An overall score of 6 with a “Below average” landing page experience is a different problem than a 6 with a “Below average” expected CTR, and the fix is different in each case.

The score also shows a dash (“—”) when Google does not have enough recent impressions to calculate it for the selected date range.^[6] Low-volume keywords frequently show dashes; that does not mean Quality Score is zero — it means there is insufficient data to report a number. Historical Quality Score columns record the last known score in a date range, and Google has provided daily historical data back to January 22, 2016.^[13]

To see these columns, you need to add them manually. They are not shown by default.

When to pull it, and when not to

The Quality Score column is useful as a triage tool, not a performance target.

• What it changes: Adding the columns gives you a per-keyword breakdown of which quality component is weakest, letting you direct copy or landing-page work to the keywords where it matters most.
• Why it works: Each component status is a relative comparison against competitors on that keyword. A “Below average” flag means your ad or page is objectively weaker than what most advertisers show for the same query, which typically means the auction is pricing you higher than a better-quality advertiser would pay for the same position.
• When to use it: Run a keyword-level Quality Score audit when your Search campaigns show an impression-weighted average CPC more than 20% above industry benchmarks for your category, or when the Auction Insights report shows you are being outranked on more than 40% of auctions. Those signals suggest a quality gap, not just a bid gap.
• When not to use it: Do not optimize toward the 1-to-10 number as a primary KPI. If your campaign’s Cost per Conversion is meeting its target and impression share is healthy, chasing a Quality Score of 10 on mid-volume keywords wastes time that belongs on budget, bidding strategy, or audience work. Quality Score is a diagnostic; the auction uses auction-time quality, which can diverge from the reported number. Also, do not rely on the score for any keyword showing a dash — there is not enough data to act on, and changes to that keyword based on a missing score are guesses.

11. Smart Bidding and the Auction

Smart Bidding is Google’s family of automated bid strategies — including Maximize Conversions, Maximize Conversion Value, Target CPA (cost per acquisition — the average amount you want to pay for a conversion), and Target ROAS (return on ad spend — the revenue you want per dollar spent on ads) — that use machine learning to set a bid for every individual auction rather than relying on a static max CPC you manage yourself.^[16]

The key distinction is where the bid is set. With manual CPC, you set a ceiling before the auction runs and it applies to every matching query until you change it. With Smart Bidding, Google calculates a custom bid milliseconds before each auction using a wide range of real-time signals: device, location, time of day, audience list membership, language, operating system, and additional signal combinations that Google does not fully disclose.^[16] The bid that enters the auction for a Chicago mobile searcher at 7 p.m. on a Tuesday can be materially different from the bid for a desktop searcher in the same city at 9 a.m., even for the same keyword — because the system’s predicted conversion probability differs between those two contexts.

This is precisely why most manual bid adjustments are ignored or limited when Smart Bidding is active. Google’s documentation says that setting a manual device, location, or schedule adjustment on a Smart Bidding campaign is redundant at best, because the algorithm is already evaluating those dimensions — and their combinations — at the per-auction level.^[16] A blunt “+20% on mobile” instruction conflicts with a model that is already deciding the mobile bid from first principles for each search. The advertiser controls you retain under Smart Bidding are: the performance target (Target CPA dollar amount or Target ROAS percentage), the daily budget, seasonality adjustments for short promotions (Google recommends 1 to 7 days), and data exclusions to block corrupted conversion periods from training the model.^[16]

There is an important honesty caveat here. Google publishes the categories of signals Smart Bidding uses, but it does not publish the exact weighting, coefficients, or per-auction formula.^[16] You can confirm which signals the strategy considers; you cannot audit the math behind a specific bid. That means you are delegating the bid decision to a model you cannot fully inspect, in exchange for per-auction precision you could never achieve manually.

What Smart Bidding does not change is the auction itself. Ad Rank still governs position and price. Smart Bidding changes the bid that enters the auction; it does not change the rules by which that bid competes. An account with poor Quality Score running Smart Bidding still pays more per click than a competitor with a higher Quality Score and a lower bid — because the auction’s pricing mechanism, shown in section 9, applies regardless of how the bid was generated.

Note that as of August 17, 2026, Google updated how Smart Bidding handles budget-constrained campaigns: campaigns limited by budget on Target CPA or Target ROAS now optimize more literally toward the target and flex less to capture additional volume.^[17][18] If your campaign is marked as “Limited by budget” in the status column, your CPA and ROAS targets need to be aligned more tightly with the budget, or the strategy will under-deliver.

12. AI Max and Where the Auction Is Going

AI Max for Search campaigns is an optimization layer — not a new campaign type — that Google made generally available in April 2026 and began automatically migrating eligible campaigns into starting September 1, 2026.^[20][21] It has three core features: search term matching expansion (the campaign can enter auctions for queries beyond your exact keyword list, using AI-driven query interpretation), text customization (headlines and descriptions can be adapted from your existing assets and landing-page content to match the specific query), and final URL expansion (Google can send the user to a more relevant page on your site rather than the URL you specified).^[22]

The migration context matters. Dynamic Search Ads (DSA) — the previous method for automatically targeting queries based on website content — cannot be created in new campaigns as of the September 2026 rollout. Existing Search campaigns using DSA, automatically created assets, or campaign-level broad match are being auto-upgraded into AI Max during September 2026.^[20][21] If you have a campaign with any of those settings and have not reviewed it since August 2026, check its status now. For a full explanation of the DSA retirement and what it means for your targeting, see the Omologist article on Dynamic Search Ads.

What AI Max does not change is the auction’s core mechanics. Ad Rank still determines which eligible ad wins each auction and what price is charged. AI Max changes which searches your campaign becomes eligible to enter, which creative variant competes, and which landing page the click leads to — but once the campaign is matched to a query, the same six-factor Ad Rank system from section 4 applies.^[22] In practical terms, AI Max increases the number of auction opportunities by expanding query coverage, and it can improve per-query relevance through customized creative. Advertisers should expect more impressions on non-obvious queries and potentially different CPC and CPA dynamics as the system makes more of the matching and message decisions before the auction.

Ads in AI Overviews and AI Mode are the next surface where auction eligibility is expanding. As of August 17, 2026, Google documented that text and Shopping ads from existing Search, Shopping, and Performance Max campaigns are eligible to appear within AI Overviews when they win the auction and match the query and overview content.^[23] No new campaign type is required; the same campaigns already running can qualify for these placements. Treat AI Overviews as additional search inventory, not a separate system requiring a separate campaign structure.

What stays true regardless of how AI Max, query expansion, or AI placements evolve: ad quality, landing-page relevance, and bid efficiency remain the variables that determine whether a campaign wins auctions cheaply or expensively. A campaign with poor Quality Score inputs entering more auctions through AI Max will enter more expensive auctions, not cheaper ones. AI changes the eligibility surface; it does not override the pricing mechanism.

To find and review the AI Max controls and the search terms it has added to your campaign, follow these steps.

13. Auction Myths

Every auction system attracts folklore, and Google Ads is no exception. The myths below persist because they contain a grain of truth — or because they are impossible to disprove with a single account’s data. The table summarizes what the evidence actually shows, followed by a plain-language explanation of each myth and why it sticks.

Myth 1: Ad Strength affects Ad Rank

Ad Strength — the “Poor / Good / Excellent” label shown inside a responsive search ad — is a creative guidance tool. It tells you how much combination variety your headlines and descriptions provide. Google’s six stated Ad Rank factors are bid, auction-time ad quality, Ad Rank thresholds, auction competitiveness, context, and expected impact of assets and formats.^[1] Ad Strength does not appear on that list, and Optmyzr’s study of more than one million ad instances found no clear correlation between Ad Strength and CTR or conversion rate.^[7] A separate 2026 RSA study reached the same conclusion.^[8] The confusion is understandable: both Ad Strength and Quality Score relate to “ad quality,” but they measure different things for different purposes. Ad Strength is a setup prompt, not an auction signal. For more on Ad Strength, see the Omologist article on Ad Strength.

Myth 2: Pausing a keyword or campaign resets Quality Score

This myth probably originates from the observation that a paused keyword sometimes shows a dash (“—”) where the Quality Score column should be. That dash means Google does not have enough recent data to display a current score — not that the score was erased. When the keyword or campaign is unpaused and impressions accumulate, the score reappears, typically close to where it was before. Tracking tools that record Quality Score daily confirm that scores persist through paused periods rather than resetting to zero.^[24][25] The practical implication: do not unpause a keyword purely to “protect” its Quality Score. The score was not in danger.

Myth 3: Spending more money improves Quality Score

Quality Score is built from three components — expected CTR, ad relevance, and landing page experience — and each is a relevance signal, not a spend signal.^[6] A $50-a-day campaign with tight ad groups, a well-matched landing page, and a compelling offer can outperform a $5,000-a-day campaign with sloppy structure. The mechanism is straightforward: Google is estimating whether your ad deserves a click for a given query, not rewarding you for your investment. Spend more if you want more auction entries; it will not raise the score in those auctions.

Myth 4: Google penalizes low budgets in the auction

A limited budget does reduce your visibility — specifically, it limits the number of auctions your campaign enters before the daily budget is exhausted, which Google reports as impression share lost to budget. But that is a reach problem, not a quality problem. In the auctions your campaign does enter, your Ad Rank is calculated the same way as a campaign with a $10,000 daily budget. Budget is not an Ad Rank input under any published Google documentation.^[1][10] If your impression share lost to budget is above 20%, the correct fix is to raise the budget or tighten keyword targeting, not to worry about an auction penalty.

Myth 5: A higher bid guarantees a higher position

Return to the four advertisers introduced in section 3. Advertiser B bids $6.00 — the highest of the four — yet finishes third. Advertiser A bids $4.00 and finishes first because Ad Rank 32 (= $4.00 × QS 8, simplified) beats Ad Rank 24 (= $6.00 × QS 4, simplified). The auction rewards the combination of bid and quality; neither alone determines position.^[3] This is the single most important structural fact in the whole article, and the reason quality work has a financial return: a higher Quality Score lets you buy the same position for less, or a better position for the same spend.

14. Reading Auction Insights Against This Model

The Auction Insights report is Google’s closest thing to a window into the auctions your ads are actually entering. It does not show individual competitor bids or Quality Scores — Google does not disclose those — but it does show five metrics that become immediately interpretable once you understand Ad Rank. Each metric is an auction outcome, not an abstract number.

Impression share — the percentage of auctions you were eligible to enter in which your ad actually showed — is your coverage rate. A benchmark range of 60% to 80% is cited for competitive accounts.^[26] If your impression share is 45%, you are absent from more than half the auctions you could have entered, and the two named causes are budget and Ad Rank. Google reports them separately: “impression share lost to budget” and “impression share lost to rank.”

Overlap rate is the share of impressions where a specific competitor’s ad also appeared. A high overlap rate with a competitor means you are entering most of the same auctions; their position relative to yours depends on their Ad Rank versus yours.

Position above rate is the share of overlapping impressions where a specific competitor’s ad appeared in a higher position than yours. If Competitor X has a position above rate of 70% against you, they are winning the Ad Rank comparison in 70% of the auctions you share. The fix is either to improve your quality inputs or to raise your bid — and the cheaper fix depends on which of your QS components is “Below average.”

Top of page rate is the share of your impressions that appeared above organic results. This reflects whether your Ad Rank consistently clears the higher threshold required for top placement.^[2] If your top of page rate is low despite a reasonable impression share, your Ad Rank is winning auctions but not clearing the top-position threshold — a signal that quality or bid is holding you just below that line.

Outranking share is the share of auctions where your ad showed in a higher position than a competitor’s, or your ad showed and theirs did not. It is a combined measure: you can outrank by appearing when they do not, or by appearing above them when both show.

Read these five metrics as a system. High impression share plus low top-of-page rate means your Ad Rank clears the floor but not the top threshold — raise quality or bid. Low impression share lost to rank means quality is the bottleneck — Ad Rank is the issue, not budget. High position above rate from a competitor with a known lower budget suggests their Quality Score is materially higher than yours.

For a full treatment of how to act on each metric, see the Omologist article on Auction Insights.

When to pull it, and when not to

The Auction Insights report is most useful when you can act on what it shows. Use it when:

• Your Search impr. share lost (rank) is above 20% and you need to identify whether competitors are systematically outranking you or whether the rank loss is spread across many small competitors — the named competitor rows answer this.
• You want to know whether a specific competitor has entered or exited your auctions after a campaign or bid change — compare a 7-day Auction Insights report before and after the change.
• Your position above rate from a named competitor exceeds 60%, meaning they are consistently above you; that gap warrants a Quality Score audit before a bid increase.

Do not use Auction Insights as your primary optimization signal when your campaign has fewer than approximately 1,000 impressions in the selected date range — the competitor metrics become statistically unreliable at low volumes and can mislead. Also, do not treat a competitor’s high outranking share as proof they are bidding more; the four-advertiser model shows a lower bid with higher Quality Score produces the same outranking result.

15. What You Can Actually Influence

The auction has six inputs, but you control only a subset of them directly. The table below maps every meaningful lever to the auction input it moves, gives a realistic size of effect based on the benchmarks in this article, and points to the Omologist article that teaches the mechanic in depth. Use this table as the navigation hub for your next step after understanding the auction model.

16. FAQs

How does Google decide who is on top of the search results?

Google calculates an Ad Rank value for every eligible ad in each auction using six factors: your bid, the quality of your ad and landing page at auction time, minimum Ad Rank thresholds, how competitive that specific auction is, the context of the search (device, location, time, query wording), and the expected impact of any ad assets you have set up.^[1] The ad with the highest Ad Rank gets the top position. A lower bid with a higher Quality Score can and does beat a higher bid with a lower Quality Score — that is the central fact of the system.

Why is my CPC so high even though I set a low max bid?

Your actual CPC is determined by the auction, not by your max bid alone. If your Quality Score is low — say, a 3 or 4 out of 10 — the simplified formula shows you need a much higher bid to generate the same Ad Rank as a competitor with a Quality Score of 8. The result is that you pay close to your full max bid in most auctions, while a higher-quality competitor pays well below theirs. Keywords with Quality Score 4 carry CPCs estimated at up to 64% above the industry median.^[15] The fix is to improve ad relevance and landing page experience, not to raise the bid further.

Does a higher bid guarantee a higher position?

No. Bid is one of six Ad Rank inputs, and quality can outweigh it. In the four-advertiser example used throughout this article, Advertiser B bids $6.00 — the highest — but finishes third because its Quality Score of 4 produces a simplified Ad Rank of 24, behind Advertiser A’s Ad Rank of 32 ($4.00 bid × QS 8) and Advertiser C’s Ad Rank of 27 ($3.00 × QS 9).^[3] Raising a bid raises Ad Rank, but if a competitor has a substantially higher Quality Score, the bid increase required to overtake them may be larger than improving quality would require.

What is Quality Score and does it still matter in 2026?

Quality Score is a 1–10 keyword-level diagnostic reported in the Google Ads interface. It is built from three component statuses — expected CTR, ad relevance, and landing page experience — and Google explicitly says it is not itself the auction input; the real auction uses auction-time ad quality, which is evaluated fresh for each search.^[6] Quality Score still matters as a diagnostic: a “Below average” status on any component tells you where to focus your quality work, and that work directly improves the auction-time signals that do determine Ad Rank and actual CPC.

What does it mean when my ad is not showing at all?

Your ad can be absent from results for several reasons: your budget has been exhausted for the day, your keywords do not match the queries being entered, your Ad Rank is below the minimum threshold Google requires for any position, your campaign or ad group is paused or disapproved, or your targeting settings exclude the searcher’s location or device.^[4][5] Open the Search terms report to confirm whether your keywords are matching queries; check the Campaigns table for “Limited” or “Eligible (limited)” status flags; and check impression share lost to budget versus lost to rank to separate the budget problem from the quality problem.

How does Google Ads pricing work — do I pay my full bid every time?

You pay your actual CPC, which is almost always lower than your max CPC bid. The actual CPC is set by the auction: Google charges you the minimum amount needed to maintain your Ad Rank above the advertiser immediately below you, plus one cent (simplified).^[11] Because that amount depends on the next advertiser’s bid and Quality Score — which change from auction to auction — your actual CPC varies by query, device, time of day, and location, even for the same keyword. Your max CPC is a ceiling, not the price you pay.

Does pausing my campaign hurt my Quality Score?

No. Pausing a campaign or keyword does not reset or lower Quality Score. When the keyword has not received impressions for a period, the Quality Score column may show a dash because Google lacks enough recent data to display a current score — but that is a reporting gap, not a score reduction.^[24] When the keyword is active again and accumulates impressions, the score reappears. Do not let fear of losing Quality Score prevent you from pausing underperforming campaigns while you fix copy or landing pages.

What is impression share and how do I know if mine is too low?

Impression share is the percentage of auctions in which your ad was eligible to appear and actually did appear.^[3] If your impression share is 45%, your ad showed in 45 out of every 100 eligible auctions. A range of 60% to 80% is cited as typical for competitive accounts.^[26] To diagnose why it is low, look at two columns in the Campaigns table: Search impr. share lost (budget) and Search impr. share lost (rank). If rank loss exceeds budget loss, the bottleneck is Ad Rank — improve Quality Score inputs or raise bids. If budget loss is larger, increase the daily budget or tighten keyword targeting to extend the budget across fewer, more valuable auctions.

---

References

1. [1] https://support.google.com/google-ads/answer/1752122?hl=en support.google.com
2. [2] https://support.google.com/google-ads/answer/1722087?hl=en support.google.com
3. [3] https://support.google.com/google-ads/answer/6366577?hl=en support.google.com
4. [4] https://support.google.com/google-ads/answer/1722122?hl=en-GB support.google.com
5. [5] https://support.google.com/google-ads/answer/7634668?hl=en support.google.com
6. [6] https://support.google.com/google-ads/answer/6167118?hl=en support.google.com
7. [7] https://www.optmyzr.com/blog/google-ad-strength-study/ www.optmyzr.com
8. [8] https://www.optmyzr.com/blog/google-rsa-performance-study/ www.optmyzr.com
9. [9] https://blog.google/products/marketingplatform/analytics/ blog.google
10. [10] https://support.google.com/google-ads/answer/156066?hl=en support.google.com
11. [11] https://support.google.com/google-ads/answer/6297?hl=en support.google.com
12. [12] https://blog.google/products-and-platforms/products/ads/big-or-small-stay-close-all-latest… blog.google
13. [13] https://searchengineland.com/google-adwords-quality-score-reporting-improvements-275010 searchengineland.com
14. [14] https://metricnexus.ai/blog/google-ads-benchmarks-2026 metricnexus.ai
15. [15] https://www.digitalapplied.com/blog/google-ads-benchmarks-2026-cpc-ctr-cvr-industry www.digitalapplied.com
16. [16] https://support.google.com/google-ads/answer/7065882?hl=en support.google.com
17. [17] https://liftconversions.com/the-lift-brief-this-week-in-ai-search-and-marketing-august-14-… liftconversions.com
18. [18] https://anicca.co.uk/blog/weekly-update-search-marketing-21-08-2026/ anicca.co.uk
19. [19] https://www.searchenginejournal.com/what-is-a-good-ctr-for-google-ads/492785/ www.searchenginejournal.com
20. [20] https://blog.google/products/ads-commerce/dsa-upgrade-to-ai-max-2026/ blog.google
21. [21] https://searchengineland.com/google-sets-ai-max-migration-timeline-for-search-campaigns-48… searchengineland.com
22. [22] https://developers.google.com/google-ads/api/docs/campaigns/ai-max-for-search-campaigns/ge… developers.google.com
23. [23] https://support.google.com/google-ads/answer/16297775?hl=en support.google.com
24. [24] https://help.optmyzr.com/en/articles/3075103-quality-score-tracker-user-guide help.optmyzr.com
25. [25] https://help.optmyzr.com/en/articles/3075087-about-quality-score-tracker help.optmyzr.com
26. [26] https://www.webfx.com/blog/ppc/google-ads-benchmarks/ www.webfx.com
27. [27] https://www.traficxo.com/blog/google-click-through-rates-ctrs-by-ranking-position-in-2026 www.traficxo.com