What Google Ads Ad Strength Really Means

Ad Strength is not part of Ad Rank or Quality Score, and an Excellent rating does not mean an ad performs. What it does measure, why chasing Excellent can backfire, and how to reach a solid Good without padding.

Last updated: November 8, 2026

In This Guide

  1. Executive Summary
  2. Benchmarks & Numbers at a Glance
  3. What It Measures
  4. The Five Ratings
  5. Does It Affect Delivery?
  6. Ad Strength vs Performance
  7. Why Chasing Excellent Backfires
  8. Reaching a Solid Good
  9. Pinning
  10. RSA vs Performance Max
  11. Common Mistakes
  12. What Changed Recently
  13. References

1. Executive Summary

Ad Strength is a creative diagnostic, not a performance score. It tells you whether the assets inside a responsive search ad (RSA), Performance Max asset group, or Demand Gen ad meet Google’s structural best practices for quantity, quality, relevance, and diversity. It is rated on a five-point scale—Incomplete, Poor, Average, Good, and Excellent—and it has no direct role in Ad Rank, Quality Score, auction eligibility, or bid calculations. Google has publicly confirmed that Ad Strength is not used in Ad Rank.[1]

The evidence base as of November 2026 complicates the intuitive assumption that a higher rating means better results. A November 2026 study of 5,900 Google Ads accounts found that lower-rated RSAs produced a higher conversion rate in 55.0% of ad-group comparisons and won more conversions per 1,000 impressions in 54.3% of comparisons.[2] Google’s own vendor claim for Performance Max is that improving Ad Strength to Excellent is associated with 6% more conversions on average—a directional aggregate, not a guarantee for any individual account.[3] Where those two data points conflict, the most conservative and accurate interpretation is that the correlation between Ad Strength and real performance is weak and inconsistent, and that chasing a higher rating can actively hurt results when it means padding an ad with weak, repetitive, or off-theme assets.

The recommended target is a solid Good. Good means the ad has sufficient breadth and depth of genuinely distinct, relevant assets—enough for Google to assemble meaningful combinations—without the padding that Excellent sometimes requires. Aim for 10–15 strong, differentiated headlines and 3–4 strong descriptions for RSAs. For Performance Max, fill every required asset type with varied, on-brand creative before worrying about optional extras. For Demand Gen, prioritize multiple visual concepts across all three key aspect ratios.

Use Ad Strength as a checklist prompt, not a KPI. It is most useful for catching genuinely thin ads: too few headlines, near-duplicate copy, no keyword coverage, or excessive pinning that collapses combinations. It is least useful—and potentially misleading—as a proxy for conversion quality, profitability, or auction competitiveness.

2. Benchmarks and Numbers at a Glance

Metric Typical range or threshold Applies when Source
Higher-rated RSA had higher CTR than lower-rated RSA in the same ad group 52.5% of ad-group comparisons Study of 5,900 US and international Google Ads accounts, November 2026; not a universal guarantee [2]
Higher-rated RSA had higher conversion rate 45.0% of ad-group comparisons Same study; lower-rated RSA won in the remaining 55.0% of comparisons [2]
Lower-rated RSA had higher conversion rate 55.0% of ad-group comparisons Same study; primary finding driving the recommendation to target Good, not Excellent [2]
Lower-rated RSA won conversions per 1,000 impressions 54.3% of comparisons Same study; illustrates that impression volume does not translate to conversion efficiency [2]
Unpinned Excellent RSAs as share of all unpinned ads 33% of unpinned RSAs Adalysis study, October 2026; share of the population that actually reaches Excellent without pinning [4]
Performance Max conversion increase when Ad Strength improves from Poor to Excellent 6% more conversions on average, at similar CPA or ROAS Google vendor claim; sample size not disclosed; not a per-account guarantee [3]
Pinning 1–2 assets to any position: maximum achievable rating Good (Excellent is blocked) Adalysis / Relevant Audience study, November 2026; pinning 1–2 assets in a single position caps the rating at Good [5]
Pinning 3 or more assets to a single position: Excellent achievable? Yes—Excellent remains possible Same study; pinning a third distinct asset to the same position preserves combination flexibility enough to allow Excellent [5]
Fully pinned Poor ads vs other pinning combinations Highest conversion rates among pinning configurations tested Same study; illustrates that a structurally restricted ad can still outperform on conversions [5]
Minimum RSA headlines required (absolute minimum to serve) 3 headlines Google RSA specification; Incomplete rating until this floor is met [1]
Maximum RSA headlines allowed 15 headlines Google RSA specification; more slots do not guarantee Good or Excellent without relevance and diversity [1]
Maximum RSA descriptions allowed 4 descriptions Google RSA specification; using all 4 with distinct messaging is recommended practice [1]
Google RSA performance improvement claim (Poor to Excellent) 15% more clicks and conversions on average Google vendor claim for RSAs; aggregate directional finding, sample size not disclosed; treated as directional only [6]

3. What Ad Strength Actually Measures

Ad Strength evaluates four interlocking dimensions of a creative setup: quantity (are enough assets present?), quality (are individual assets clear, distinct, and policy-compliant?), relevance (do assets match the ad group’s keywords, intent, and landing page?), and diversity (do assets communicate meaningfully different messages, or do they repeat the same claim with minor wording changes?).[7]

For RSAs, those four dimensions are assessed across the available pool of headlines and descriptions and the combinations they can form. For Performance Max, two additional dimensions are added: breadth (does the asset group contain the formats required to serve every eligible inventory surface?) and depth (is there enough variety within each format for the system to test alternatives?).[7]

Critically, Ad Strength does not measure any of the following: historical click-through rate, actual conversion rate, actual CPA, Quality Score, Ad Rank, landing-page experience score, or search-term match quality. It is a rules-and-recommendations engine applied to the creative configuration at a point in time. Google does not publish a numerical weighting formula—there is no disclosed breakdown such as “30% relevance, 25% diversity.” The rating is a qualitative label generated from Google’s internal best-practice evaluation.[7]

The rating is also dynamic. For RSAs, it can recalculate when keywords, assets, or account structure change, and—since September 2026—when eligible Google-generated text assets are added via the AI Max text customization feature.[8] This means an RSA’s displayed rating can shift without any manual edit from the advertiser, which is a governance risk for brands with strict message controls.

Worked example (illustrative)

Diagnosing a thin RSA in a Denver HVAC account

  • Setup: A Denver HVAC account spending $6,000 per month has one RSA in its “furnace repair” ad group with 5 headlines and 2 descriptions. Three of the 5 headlines are near-duplicates: “Denver Furnace Repair,” “Furnace Repair Denver,” and “Expert Furnace Repair Denver.” The 2 descriptions cover only price and phone number. Ad Strength shows Average.
  • Numbers: 5 headlines supplied vs. the recommended 10–15; 3 of 5 headlines (60%) are near-duplicate variants of the same phrase; 2 descriptions supplied vs. the maximum 4; 0 headlines cover benefits, trust signals, or objection handlers; 0 descriptions include a call to action or guarantee.
  • Decision: Add 7 net-new headlines covering distinct message roles (e.g., “Same-Day Service Available,” “10-Year Parts Warranty,” “Licensed & Insured Technicians,” “Financing Available—0% for 12 Months,” “Serving Denver Since 2003,” “Free Diagnostic with Any Repair,” “BBB A+ Rated”) and 2 net-new descriptions (e.g., one covering the guarantee, one covering the service area). Remove the 2 near-duplicate headline variants. Target: Good.
  • Why: The Ad Strength diagnostic correctly flags insufficient quantity and diversity; each of the 7 new headlines occupies a genuinely different message role, satisfying the diversity dimension without padding, and matches the landing page’s actual offer.

4. The Five Ratings, Explained

Google uses a five-point descriptive scale. The table below summarizes what each label means and what the practical implication is for the advertiser—not for the auction.

Rating What it signals RSA implication Performance Max implication
Incomplete Minimum assets to serve are absent Fewer than 3 headlines or no descriptions; ad cannot serve One or more required asset types are missing; asset group cannot serve all inventory
Poor Only minimum required assets present; little diversity or breadth 3–4 headlines and/or only 1 description; near-duplicate copy; heavy pinning Minimum asset types present but insufficient breadth; asset group may serve only some inventory surfaces
Average Required types are present but depth or variety is lacking Adequate headline and description count but repetitive messaging or low relevance All required asset types present but not enough variation within types
Good Required types present with good variety and text diversification 8–12 distinct, relevant headlines; 3–4 strong descriptions; minimal pinning Broad asset coverage, useful variety across formats, good text diversification
Excellent Full recommended asset mix, strong variety, optional assets included Close to 15 strong headlines; 4 descriptions; sitelinks; broad message coverage All required and optional asset types present (sitelinks, horizontal logo, square image, video); maximum diversification

Two important caveats apply to this scale. First, reaching Excellent does not guarantee auction wins, clicks, or conversions—the November 2026 study found lower-rated RSAs produced higher conversion rates in 55.0% of comparisons.[2] Second, Poor does not mean the ad is penalized or throttled; Google has confirmed that Ad Strength is not used in Ad Rank for RSAs.[1] The labels describe structural completeness, not commercial effectiveness.

Worked example (illustrative)

What keeps an RSA at Average despite 12 headlines

  • Setup: A Chicago personal injury law firm spending $18,000 per month has an RSA with 12 headlines and 3 descriptions, rated Average. The firm’s compliance team requires that all headlines reference “No Fee Unless We Win.” Ten of the 12 headlines are variations of that phrase. The 3 descriptions are near-identical restatements of the contingency fee policy.
  • Numbers: 12 headlines supplied (within the recommended range); 10 of 12 headlines (83%) repeat the same contingency-fee concept; effective diversity score: approximately 2 meaningfully distinct message roles out of a recommended minimum of 5–6; 3 descriptions present but all 3 cover a single theme.
  • Decision: Keep 2–3 contingency-fee headlines (the message is legally important and differentiating) and add 7–8 headlines covering distinct roles: practice areas, local credibility, case results (where state bar rules permit), response speed, and client reviews. Add 1 net-new description covering the firm’s specific practice areas. Target rating: Good. Do not pursue Excellent if doing so requires adding claims that cannot be substantiated under Illinois Rules of Professional Conduct.
  • Why: The Average rating is driven by insufficient diversity, not insufficient quantity; Google’s diversity dimension penalizes near-duplicate assets regardless of how many total headlines are supplied.[7]

Managed for you

Everything on this page, handled every month

You are reading how to do it properly. Omologist does it, inside a written mandate, for 15% of your budget.

✓  Conversion tracking built and tested before any ad runs

✓  Search campaigns built to what your budget can actually feed

✓  Changes made inside a written mandate, not whenever

✓  A report on the 1st that says what was done and why

See the service and prices

15% of budget plus GST · minimum A$330 incl GST · accounts under about A$10,000 a month

5. Does Ad Strength Affect Whether or How Your Ads Run?

The direct answer is no, for RSAs. Google states explicitly that Ad Strength does not directly influence serving eligibility and is not used to calculate Ad Rank, Quality Score, or auction wins.[1] A Poor-rated RSA competes in every auction on the same Ad Rank factors as an Excellent-rated RSA: bid, expected CTR, ad and landing-page quality, context, and format extensions. The Ad Strength label attached to the ad is not one of those factors.[9]

For Performance Max, the relationship is more nuanced. While the rating itself is not an auction input, the underlying assets determine which inventory surfaces the asset group is eligible to serve. A Poor rating often means the asset group lacks required formats—no video, no square image, no logo—which limits physical access to YouTube, Display, or Discover inventory. In that case, the rating is a symptom of an inventory-coverage problem, not a penalty. Fixing the missing asset types expands eligibility and can increase impressions and conversions independently of any “rating improvement” effect.[7]

Google reports a 6% average conversion increase for Performance Max advertisers who improve Ad Strength to Excellent, but this is a vendor claim with no disclosed sample size or methodology.[3] The most rigorous interpretation is that adding genuinely useful assets—video, images in multiple aspect ratios, sitelinks—can unlock inventory and improve results, while adding repetitive or low-quality assets just to reach Excellent does not reliably produce the same effect.

One important 2026 development: when AI Max text customization is enabled, Google-generated headlines and descriptions can be counted in the RSA’s Ad Strength calculation. This means the rating can change without any advertiser action, and the ad pool may contain assets the advertiser did not write or approve. This is a governance issue, not a performance guarantee.[8]

Worked example (illustrative)

Performance Max asset group unlocking YouTube inventory by fixing a Poor rating

  • Setup: An Atlanta e-commerce account selling home fitness equipment spends $22,000 per month on a Performance Max campaign. The single asset group is rated Poor. The campaign runs on a Merchant Center feed but has no uploaded video, no square logo, and no lifestyle images—only product images from the feed.
  • Numbers: Current asset group contains: 5 headlines, 2 descriptions, 8 product images (feed-sourced), 0 uploaded videos, 0 square logos, 0 lifestyle images. YouTube inventory: $0 spend (0 impressions) over the prior 30-day period (October 1–October 31, 2026) because no video asset is present. Google’s vendor claim: improving Poor to Excellent is associated with 6% more conversions on average.[3]
  • Decision: Upload 1 horizontal logo (1200×628 px), 1 square logo (1200×1200 px), 3 lifestyle images (at least one each in square, landscape, and vertical format), and 1 brand video of at least 10 seconds optimized for YouTube. Add 5 net-new headlines covering benefit and lifestyle themes (distinct from the existing product-feature headlines) and 2 net-new descriptions. Target rating: Good. Do not add a second video or additional images solely to reach Excellent if the creative team cannot produce distinct, on-brand concepts within the Q4 2026 production cycle.
  • Why: The Poor rating reflects missing required asset types (video, logo), which physically blocks YouTube inventory eligibility; fixing the asset types addresses the coverage problem regardless of its effect on the displayed rating.[7]

6. Ad Strength vs Real Performance: What the Evidence Shows

The most rigorous independent data point available as of November 2026 is a study of 5,900 Google Ads accounts published by Search Engine Journal on November 3, 2026.[2] Its key findings are:

  • Higher-rated RSAs had a higher CTR in 52.5% of ad-group comparisons—a slim majority, and far from universal.[2]
  • Higher-rated RSAs had a higher conversion rate in 45.0% of comparisons, meaning lower-rated RSAs won on conversion rate in 55.0% of comparisons.[2]
  • Lower-rated RSAs produced more conversions per 1,000 impressions in 54.3% of comparisons.[2]

These figures do not prove that lower Ad Strength causes better performance—the relationship is not causal in either direction. What the data shows is that Ad Strength is a weak and inconsistent proxy for the metrics that actually matter: conversion rate, conversions per impression, and implied CPA. An ad can be structurally complete (high Ad Strength) while being strategically weak (generic headlines that generate clicks from unqualified searchers), or structurally lean (low Ad Strength) while being strategically sharp (tightly focused copy that filters to high-intent buyers).

Google’s own aggregate directional claims—15% more clicks and conversions for RSAs improving from Poor to Excellent[6] and 6% more conversions for Performance Max[3]—are not contradicted by the independent data, but they describe a population average that conceals wide variation. Both sets of numbers can be true simultaneously: on average, more diverse assets help; in a majority of individual cases, the lower-rated ad wins on conversion rate.

The practical implication is to treat Ad Strength as a creative-process input, not a performance forecast. Measure success with conversion tracking, search-term analysis, asset-level performance reports, and controlled experiments—not the displayed rating.

Worked example (illustrative)

Interpreting conflicting Ad Strength and conversion data in the same ad group

  • Setup: A Seattle B2B SaaS account spending $14,000 per month on Search has two RSAs in its “project management software” ad group. RSA-A is rated Excellent (14 headlines, 4 descriptions, all unpinned). RSA-B is rated Good (10 headlines, 4 descriptions, 1 headline pinned to Position 1 for the brand name). Both ads have been running since January 1, 2026.
  • Numbers: Over the 90-day period July 1–September 30, 2026: RSA-A CTR = 4.8%, RSA-B CTR = 4.1%. RSA-A conversion rate = 3.2% (128 conversions from 4,000 clicks); RSA-B conversion rate = 4.9% (147 conversions from 3,000 clicks). RSA-A CPA = $109.38 (($14,000 × RSA-A share of spend) / 128); RSA-B CPA = $95.24 (illustrative). The pattern matches the November 2026 study finding: lower-rated RSA won on conversion rate in 55.0% of comparisons.[2]
  • Decision: Do not restructure RSA-B to reach Excellent. Instead, run a Google Ads experiment (50/50 traffic split, minimum 4-week duration, minimum 100 conversions per arm) to test whether adding 4 more headlines to RSA-B improves conversion rate without raising CPA above $110.
  • Why: CTR advantage for the higher-rated ad does not translate to conversion advantage; the operative KPI for this B2B account is CPA, and Ad Strength does not predict CPA.[2]

Not ready to change anything today?

We post the sharpest bits of these guides — what a change is worth, how long it takes to show up, and when to leave it alone. Short enough to read on your phone.

Follow on LinkedIn Follow on Instagram

7. Why Chasing Excellent Can Backfire

The path from Good to Excellent for an RSA typically requires pushing headline count toward 15 and description count to 4, with strong diversity across all assets. For many real ad groups, the genuinely distinct, accurate, and relevant messages run out before slot 15. What fills the remaining slots tends to be one or more of the following failure modes:

  • Generic filler headlines (“Quality Service,” “Get Started Today,” “Learn More Now”) that add nothing differentiating and can combine with specific product headlines to produce incoherent ad variations.
  • Near-duplicate variants (“Fast Shipping,” “Quick Shipping,” “Ships Fast”) that superficially satisfy the diversity check but communicate identical propositions.
  • Off-theme assets that technically belong to the brand but are mismatched with the ad group’s keyword intent, creating combinations that confuse searchers.
  • Weak calls to action added purely to reach a slot count (“Visit Our Website,” “Contact Us Now”) without a compelling reason to act.
  • Unsubstantiated claims that an advertiser’s legal or compliance team would not approve if reviewed—introduced because the creative brief says “get to Excellent.”

Each of these failure modes can lower the average quality of the combinations Google assembles, reduce click-to-conversion rates, and in some cases introduce regulatory or brand-governance risk. The November 2026 study data supports this concern: lower-rated RSAs outperformed on conversion rate in 55.0% of comparisons, which is consistent with the hypothesis that tighter, more relevant copy—even if it means fewer assets and a lower rating—outperforms padded Excellent ads for bottom-funnel efficiency.[2]

The pinning data adds another dimension. The Adalysis/Relevant Audience study from November 2026 found that fully pinned Poor ads produced the highest conversion rates among all pinning configurations tested.[5] This is a specific, narrow finding—not a recommendation to build Poor ads—but it confirms that structural control over ad assembly can outperform algorithmic combination freedom in some contexts, particularly for advertisers with strong conversion data, precise intent signals, or legally sensitive messaging.

Worked example (illustrative)

Padding an RSA to Excellent and measuring the downstream cost

  • Setup: A Minneapolis mortgage broker account spending $9,500 per month improves an RSA from Good (11 headlines, 4 descriptions) to Excellent (15 headlines, 4 descriptions) by adding 4 headlines: “Get Your Quote Online,” “We’re Here to Help,” “Start Your Application,” and “Mortgage Solutions Available.” The compliance team approved the additions in under 10 minutes because they contained no rate claims. Ad Strength moves from Good to Excellent.
  • Numbers: Pre-change (August 1–September 30, 2026, 60 days): CTR = 5.2%, conversion rate = 6.8%, qualified leads = 82, CPA = $231.71. Post-change (October 1–November 30, 2026, 60 days): CTR = 5.6% (+0.4 percentage points), conversion rate = 5.1% (−1.7 percentage points), qualified leads = 71, CPA = $267.61 (+$35.90 per lead, +15.5%). The 4 added headlines are generic across the mortgage category and generate clicks from early-stage researchers who do not convert at the same rate as intent-matched searchers.
  • Decision: Remove the 4 generic headlines and return the RSA to 11 headlines (Good rating). Brief the creative team to develop 2–3 additional benefit-specific headlines (e.g., “Lock Your Rate for 90 Days,” “Local Lender—Close in 21 Days”) only if they can be substantiated and approved by the compliance team.
  • Why: The 4 added headlines satisfied the diversity dimension of Ad Strength but reduced message specificity, attracting lower-intent clicks and raising CPA by $35.90; padding to Excellent produced a measurable negative outcome.[2]

8. How to Reach a Solid Good Without Padding

A Good Ad Strength rating for an RSA is achievable with 10–12 genuinely distinct, relevant headlines and 3–4 strong descriptions. The operative word is “genuinely distinct”: each headline should occupy a different message role, not rephrase an existing one. Google recommends including at least 5 unique headlines and at least 3 headlines that do not contain the keyword and instead cover benefits, features, shipping, returns, or other useful information.[10]

A practical message-role framework for RSAs covers eight categories:

  • Keyword/product category (1–2 headlines): The core search intent phrase, used naturally.
  • Brand (1 headline): The company name or tagline, where it aids trust or recall.
  • Primary benefit (1–2 headlines): The single strongest reason a searcher should choose this advertiser.
  • Feature or specification (1–2 headlines): A concrete, differentiating product or service attribute.
  • Price, offer, or availability (1–2 headlines): A specific dollar amount, percentage off, or inventory/timing signal.
  • Trust signal (1 headline): A review count, rating, certification, years in business, or warranty.
  • Objection handler (1 headline): Free shipping, free returns, no-contract cancellation, or a similar friction-reducer.
  • Call to action (1 headline): A specific, benefit-linked action (“Get a Free Estimate in 60 Seconds”), not a generic verb.

Eight roles, each filled with 1–2 headlines, produces 8–16 headlines. If a role cannot be filled with an accurate, substantiated claim for this ad group, skip it rather than inventing something. For descriptions, each of the 4 slots should cover a different dimension: the primary value proposition, a specific proof point, an objection handler, and the CTA with urgency or exclusivity if applicable.

Length variety also matters. Google’s guidance recommends a mix of shorter and longer headlines rather than making every asset the same length.[11] A mix of 3–5 word headlines and 8–10 word headlines helps the system assemble varied combinations across different ad formats and placements.

Worked example (illustrative)

Building a Good-rated RSA from scratch for a Portland roofing company

  • Setup: A Portland residential roofing company spending $4,800 per month is launching its first RSA in a “roof replacement” ad group targeting Portland metro. Budget is $160 per day. The landing page covers asphalt shingle and metal roof replacement with a free estimate offer and a 10-year workmanship warranty.
  • Numbers: Target: 11 distinct headlines across 8 message roles; 4 descriptions across 4 distinct dimensions. Headlines: (1) “Portland Roof Replacement” [keyword], (2) “Metal & Shingle Roofs” [product], (3) “Pacific Roofing Co.” [brand], (4) “Free Estimates—No Obligation” [offer], (5) “10-Year Workmanship Warranty” [trust], (6) “Hail & Wind Damage Specialists” [feature], (7) “Licensed Oregon Contractor” [trust], (8) “Same-Week Installation Available” [availability], (9) “Financing From $0 Down” [price/offer], (10) “1,200+ Portland Homeowners Served” [trust], (11) “Get Your Free Estimate Today” [CTA]. Descriptions: (1) full value proposition with warranty; (2) process/speed proof point; (3) financing objection handler; (4) CTA with geographic urgency. Projected Ad Strength: Good. No headline repeats a concept from another; 0 of 11 are generic filler.
  • Decision: Launch with 11 headlines and 4 descriptions targeting Good. Do not add a 12th or 13th headline unless a genuinely new message role (e.g., seasonal promotion, new service area) can be identified in Q1 2027.
  • Why: All 11 headlines cover a distinct message role aligned with the landing page; Google’s diversity dimension is satisfied without padding, and the Good rating reflects genuine creative coverage rather than quantity inflation.[7]

Worked example (illustrative)

Using length variety to improve Ad Strength without adding new concepts

  • Setup: A San Diego estate planning law firm spending $7,200 per month has an RSA rated Average. It has 9 headlines, all between 25 and 30 characters. Google’s interface shows “Add more distinct headlines” and “Use longer headlines.” No new message roles are available without adding claims the firm’s bar compliance review would not approve.
  • Numbers: Current headline character lengths (approximate): all 9 headlines fall in the 20–28 character range. Google’s RSA allows up to 30 characters per headline. The issue is not length ceiling but uniformity. Available unused character budget across all 9 headlines: average 6 characters per headline, total 54 characters unused. Action: Expand 3 existing headlines from their current short form to a longer form that adds a specific detail without changing the message role. Example: “Free Consultation” (18 chars) → “Free 30-Minute Phone Consultation” (33 chars—requires trimming to 30) → “Free Consult—Call Today” (23 chars) [not ideal]; or add 2 net-new headlines in the 25–30 character range covering a specific new detail: “Revocable & Irrevocable Trusts” (30 chars) and “Probate Avoidance Planning” (26 chars).
  • Decision: Add 2 net-new headlines covering specific practice sub-areas (“Revocable & Irrevocable Trusts” and “Probate Avoidance Planning”) that are accurate, approved, and cover a genuinely distinct message role (service specification). Do not expand headline lengths artificially. Target: Good with 11 total headlines.
  • Why: The Average rating is driven by insufficient diversity and the absence of longer, more specific headlines; adding 2 on-topic specification headlines satisfies the diversity dimension without introducing generic filler.[7]

“So what do I actually have to do?”

Tell us the business and give us the account. We build the conversion tracking and prove it works with a test conversion before any ad runs, then launch the campaigns your budget can feed.

After that, changes happen inside a written mandate, so you know in advance what gets changed without asking and what never gets touched. A report on the 1st, a budget email around the 23rd, and nothing else unless it needs your yes or no.

How it works

9. Pinning and Its Effect on Ad Strength

Pinning locks a specific headline or description to a specific position in the ad. It is a useful tool for mandatory disclosures, regulated messaging, or controlled creative tests—and a significant constraint on Ad Strength if used excessively.

The Adalysis/Relevant Audience study from November 2026 found that pinning 1–2 assets to any single position caps Ad Strength at a maximum of Good, blocking Excellent.[5] Pinning 3 or more distinct assets to the same position preserves the system’s ability to rotate between them and allows Excellent to remain achievable.[5] This creates a practical rule: if you must pin, pin a slot with at least 3 meaningfully different assets rather than 1 or 2.[7]

The same study found that fully pinned Poor ads produced the highest conversion rates among all pinning configurations tested.[5] This counterintuitive result likely reflects selection effects: advertisers who fully pin their ads tend to have strong, tested messaging with high intent alignment, often in categories where precision matters more than combination breadth (financial services, healthcare, legal, e-commerce with specific promotional constraints). The finding does not mean advertiser should build Poor-rated ads deliberately; it means that pinning-driven Ad Strength reduction does not automatically produce inferior conversion results.

The practical pinning hierarchy for 2026 is:

  • Default: no pinning. Leave all positions open to maximize combination flexibility and Ad Strength potential.
  • If one message must always appear: Pin it to its required position and add 2 additional distinct assets to the same pin slot so the system can rotate.
  • If multiple messages must always appear: Pin the essential ones; accept a Good ceiling and measure performance independently rather than chasing Excellent.
  • If the ad is in a regulated category (financial, healthcare, legal, pharmaceutical): Pin required disclosures; run controlled experiments to verify whether the pinning constraint meaningfully affects CPA before modifying.

Worked example (illustrative)

Pinning a mandatory disclosure without blocking Good

  • Setup: A Dallas auto insurance account spending $31,000 per month is required by its legal team to show the phrase “Licensed Texas Insurance Agency” in Headline Position 1 of every ad, to comply with Texas Department of Insurance advertising requirements. The RSA currently pins only that one headline to Position 1 and has 10 unpinned headlines and 4 descriptions elsewhere. Ad Strength is currently Average because pinning 1 asset to Position 1 restricts combinations.
  • Numbers: Current setup: 1 pinned headline (Position 1), 10 unpinned headlines across Positions 2–3 rotation, 4 descriptions. Per Adalysis/Relevant Audience study, pinning 1–2 assets to a position caps the rating at Good; pinning exactly 1 gives Average or Good depending on overall diversity.[5] Fix: Add 2 additional legally approved Position 1 variants to the pin slot: “Texas-Licensed Auto Insurance” and “TX Insurance Agency—Licensed & Bonded.” Now 3 assets are pinned to Position 1; Excellent becomes achievable per the study.[5]
  • Decision: Add “Texas-Licensed Auto Insurance” and “TX Insurance Agency—Licensed & Bonded” to the Position 1 pin slot (total 3 pinned to Position 1). Submit both variants for legal review by November 15, 2026. Target: Good or Excellent without removing the mandatory disclosure.
  • Why: Pinning 3 or more assets to a single position preserves combination flexibility and allows Excellent; pinning fewer than 3 caps the rating at Good at best, per the November 2026 study.[5]

10. Ad Strength in Responsive Search Ads vs Performance Max

Ad Strength applies to both RSAs and Performance Max asset groups, but the inputs, the failure modes, and the recommended remediation differ substantially between the two formats.[7]

For RSAs, Ad Strength is predominantly a text-asset diagnostic. The system evaluates whether the pool of headlines and descriptions is large enough, diverse enough, and relevant enough for Google to assemble a range of meaningful search ad combinations. The core risk is a text-only failure: too few headlines, near-duplicate copy, or heavy pinning.[1]

For Performance Max, Ad Strength is a multimedia diagnostic. It evaluates whether the asset group covers the full range of formats required to serve eligible inventory surfaces—Search, Display, YouTube, Discover, Gmail, Maps—and whether there is enough creative variety within each format. A Performance Max asset group rated Poor often reflects a missing asset type (no video, no logo, no images beyond the product feed) rather than a text problem. Fixing the missing asset types is the first priority, because doing so directly expands inventory eligibility—a concrete, measurable effect independent of the rating label itself.[7]

For Demand Gen, Ad Strength is primarily a visual-asset diagnostic, evaluating breadth across aspect ratios (square, landscape, vertical), creative diversity (image vs video), and placement coverage across YouTube, Shorts, Discover, Gmail, and Display. Google recommends at least three unique assets for each major aspect ratio.[12] This is a different creative challenge from RSAs: the bottleneck is usually production capacity for video and varied visual concepts, not copywriting bandwidth.

Dimension RSA Performance Max Demand Gen
Primary asset type evaluated Headlines and descriptions Full multimedia asset set (text, images, video, logo) Visual assets (image and video) plus text
Key breadth requirement Minimum 3 headlines; recommended 10–15 All required formats for eligible inventory surfaces All 3 key aspect ratios (square, landscape, vertical)
Most common Poor/Average cause Too few headlines, near-duplicate copy, heavy pinning Missing video, logo, or image format Missing aspect ratio, no video, insufficient visual variety
Primary fix for Poor Add distinct headlines/descriptions in new message roles Upload missing required asset types first Produce assets in missing aspect ratios; add video
Excellent requirement (notable) Close to 15 headlines; sitelinks; 4 descriptions Optional assets: sitelinks, horizontal logo, square image, video 3+ unique assets per aspect ratio; video included
Google’s average conversion claim for Excellent 15% more clicks and conversions (vendor claim)[6] 6% more conversions on average (vendor claim)[3] Not separately quantified in available research

Worked example (illustrative)

Prioritizing asset-type gaps in a Performance Max asset group before fixing text

  • Setup: A Boston direct-to-consumer supplement brand spending $27,000 per month on Performance Max has a single asset group rated Poor. The asset group has 8 headlines, 3 descriptions, 6 product images (all landscape format), 1 square logo, and 0 uploaded videos. Google’s asset group editor shows YouTube as an ineligible inventory surface because no video is present.
  • Numbers: Missing asset types: (1) vertical images (9:16 aspect ratio), (2) square images (1:1 aspect ratio), (3) video. YouTube spend over October 1–October 31, 2026: $0 (0% of $27,000 monthly budget). Estimated YouTube share of Performance Max inventory for DTC e-commerce in the US: not published, but YouTube is a primary surface; $0 from that surface represents a material coverage gap. Google vendor claim: improving Poor to Excellent → 6% more conversions on average.[3] Even if only 3% of that improvement is attributable to YouTube inventory access, at a 2.5% current conversion rate and $45 average order value, 3% more conversions from a $27,000 monthly budget represents approximately 18 additional conversions per month.
  • Decision: Week 1: Produce and upload 1 brand video (minimum 10 seconds, optimized for YouTube skippable in-stream, 16:9 aspect ratio). Week 2: Shoot 3 square (1:1) and 3 vertical (9:16) lifestyle images. Week 3: Add 4 net-new headlines in distinct message roles (ingredient benefit, clinically studied claim if substantiated, subscription savings, guarantee). Do not add a second video or additional images in Week 1–2 solely to reach Excellent.
  • Why: Missing asset types (video, vertical images) directly block inventory surface eligibility; fixing breadth gaps has a concrete, measurable effect on impression coverage that is independent of the rating label improvement.[7]

11. Common Mistakes to Avoid

The most damaging mistakes fall into three categories: misunderstanding what Ad Strength measures, over-optimizing for the rating, and under-using it as a legitimate diagnostic. The following are the most frequently observed errors in 2026 account audits.

Treating Ad Strength as a performance score

Ad Strength is not a Quality Score, not an Ad Rank input, and not a predictor of CTR or conversion rate. The November 2026 study found that lower-rated RSAs produced a higher conversion rate in 55.0% of comparisons.[2] Advertisers who set “Excellent Ad Strength across all RSAs” as a campaign KPI are optimizing against a metric that does not reliably predict business outcomes.

Adding generic headlines to reach Excellent

The most common path from Good to Excellent is adding 3–4 more headlines. If the ad group has exhausted its relevant, substantiated messages at headline 11 or 12, the remaining slots will be filled with generic copy. Generic copy lowers the average relevance of combinations Google assembles and can increase clicks from unqualified searchers, raising CPA. This is the core failure mode documented in the evidence.[2]

Ignoring Ad Strength entirely

The opposite error is dismissing Ad Strength as meaningless because it does not affect Ad Rank. For genuinely thin ads—3 headlines and 1 description, or 5 near-duplicate headlines—Ad Strength correctly identifies a real creative problem. An ad with 3 headlines and 1 description has a narrow combination set that limits Google’s ability to match messaging to search context. Fixing that is a legitimate improvement.[1]

Failing to audit AI-generated assets after enabling text customization

Since September 2026, AI Max text customization can add Google-generated headlines and descriptions to an RSA’s asset pool and count them toward Ad Strength. Advertisers who enable this feature without reviewing generated assets risk serving headlines containing incorrect pricing, unauthorized promotional claims, trademark violations, or messaging that conflicts with legal and compliance requirements.[8] Review every generated asset against the same standards applied to human-written copy.

Excessive pinning without compensation

Pinning 1–2 assets to a position without adding additional assets to that pin slot caps Ad Strength at Good and reduces combination flexibility. If pinning is required, populate the pin slot with at least 3 distinct assets to preserve rotation and avoid an artificially depressed rating.[5]

Using Ad Strength as the sole creative evaluation tool

Ad Strength does not report on individual asset performance. Google’s asset-level performance labels (Best, Good, Low, Learning) in the RSA asset report provide more granular signal about which specific headlines and descriptions are driving results. Reviewing both the Ad Strength diagnostic and the asset-level performance report gives a fuller picture than either alone.

Worked example (illustrative)

Catching an AI-generated asset governance failure after enabling text customization

  • Setup: A Phoenix real estate brokerage account spending $11,000 per month enables AI Max text customization in October 2026. Within 72 hours, the RSA’s Ad Strength improves from Average to Good as Google adds 3 generated headlines to the pool. The account manager does not review the generated assets immediately.
  • Numbers: Generated headlines added by Google: (1) “Phoenix Homes From $280,000” [contains a price claim; the brokerage’s current listing minimum is $420,000], (2) “Top-Rated Phoenix Realtor” [contains a superlative claim that may require substantiation under Arizona Department of Real Estate advertising rules], (3) “Zero Commission Sales Available” [the brokerage does not offer zero-commission listings]. All 3 headlines are factually inaccurate for this brokerage. The rating improvement from Average to Good is driven in part by assets the brokerage cannot legally or factually serve.
  • Decision: Immediately open the asset report and mark all 3 generated headlines as “Remove.” Set a governance calendar reminder to audit AI-generated assets every 7 days. Brief the compliance team by November 1, 2026 on the text customization feature and establish a written review protocol before re-enabling it.
  • Why: Google-generated text assets count toward Ad Strength when text customization is active but are not guaranteed to be accurate; the brokerage is responsible for all content served under its account, regardless of generation source.[8]

12. What Changed Recently (Last 30 Days)

As of November 2026, the most significant recent development is the full integration of AI Max text customization into RSA Ad Strength calculations, following the September 2026 rollout that began upgrading accounts from the legacy “automatically created assets” feature.[8] This changes the practical meaning of the rating in three ways:

  • The rating can now improve without advertiser action. When Google adds generated headlines and descriptions to an RSA’s asset pool, the Ad Strength recalculates automatically. An account manager reviewing a campaign weekly may see a rating change between sessions that reflects Google’s additions rather than any human edit.[8]
  • The asset pool is larger and more dynamic. Generated assets may be refreshed within 48 hours when a landing page changes. Google may also remove an asset if it receives a “Low performance” label or if its source changes.[8] This introduces volatility into the ad pool that did not exist under purely advertiser-managed RSAs.
  • English-language eligibility limitation. As of November 2026, Google-customized generative text assets are eligible only for English-language assets. Accounts targeting Spanish, French, or other languages in the US will not see the same AI-generated asset additions to their asset pools.[8]

A November 3, 2026 Search Engine Journal analysis of 5,900 accounts—the largest independent study available at this writing—confirmed that Google’s serving algorithm tends to favor higher-rated RSAs for impressions, but that lower-rated RSAs produce higher conversion rates in a majority of comparisons.[2] This finding reinforces the editorial position that Ad Strength is a diagnostic of structural completeness, not a measure of commercial effectiveness.

The Adalysis/Relevant Audience pinning study, published November 4, 2026, clarified the relationship between pinning and rating ceiling: pinning fewer than 3 assets to a position blocks Excellent; pinning 3 or more preserves it; and fully pinned Poor ads produced the highest conversion rates among the pinning configurations tested.[5] This is actionable guidance for accounts in regulated categories that require fixed messaging.

Google’s Performance Max Ad Strength guidance, last updated October 30, 2026, continues to cite the 6% average conversion increase for advertisers who reach Excellent, without disclosing sample size or methodology.[3] Treat this as a directional vendor claim, not a guaranteed outcome.

No changes to Ad Rank have been announced that incorporate Ad Strength. The rating remains a creative diagnostic tool with no confirmed role in auction mechanics as of November 2026.[1]

Worked example (illustrative)

Responding to an unexpected Ad Strength change driven by AI Max text customization in November 2026

  • Setup: A Nashville healthcare staffing firm spending $8,400 per month notices on November 10, 2026, that its primary RSA’s Ad Strength improved from Poor to Good overnight without any manual edits. The account has AI Max text customization enabled. The campaign targets travel nurse job seekers in the continental US.
  • Numbers: Prior asset count (November 9, 2026): 6 advertiser-authored headlines, 2 descriptions. Post-change asset count (November 10, 2026): 6 advertiser-authored + 4 Google-generated headlines = 10 total; 2 advertiser-authored + 1 Google-generated description = 3 total. Rating change: Poor → Good. Generated headlines include: “Competitive Travel Nurse Pay” [accurate and acceptable], “Top Travel Nurse Agency” [superlative—requires substantiation under Tennessee healthcare advertising norms and potentially FTC endorsement guidelines], “Nurse Jobs Across the US” [accurate and acceptable], “Apply for Travel Nurse Roles” [generic CTA—acceptable but weak]. 1 of 4 generated headlines carries a compliance risk.
  • Decision: By November 12, 2026: (1) Remove “Top Travel Nurse Agency” from the generated asset pool via the asset report. (2) Replace it with an advertiser-authored headline: “Joint Commission–Certified Staffing” (if the firm holds that certification). (3) Schedule a weekly AI-generated asset review every Monday at 9:00 a.m. CST. (4) Brief the HR compliance team by November 15, 2026 on the text customization feature; document the review protocol in writing.
  • Why: AI Max text customization assets count toward Ad Strength and are served under the advertiser’s account; superlative claims require substantiation under FTC guidelines and applicable state rules, regardless of whether Google or the advertiser authored the asset.[8]

References

  1. [1] https://support.google.com/google-ads/answer/9921843?hl=en support.google.com
  2. [2] https://www.searchenginejournal.com/google-favors-high-ad-strength-rsas-but-lower-rated-ad… www.searchenginejournal.com
  3. [3] https://support.google.com/google-ads/answer/14143250?hl=en support.google.com
  4. [4] https://adalysis.com/blog/google-ads-data-study-does-a-higher-ad-strength-improve-your-per… adalysis.com
  5. [5] https://www.relevantaudience.com/google-ads-en/adalysis-ad-strength-study-google-ads-2026/ www.relevantaudience.com
  6. [6] https://support.google.com/google-ads/answer/6167101?hl=en support.google.com
  7. [7] https://support.google.com/google-ads/answer/9142254? support.google.com
  8. [8] https://support.google.com/google-ads/answer/6238826?hl=en support.google.com
  9. [9] https://support.google.com/google-ads/answer/1752122?hl=en support.google.com
  10. [10] https://support.google.com/google-ads/answer/9921844?hl=en&ref_topic=10289674 support.google.com
  11. [11] https://support.google.com/google-ads/answer/9781208?hl=en-BD support.google.com
  12. [12] https://support.google.com/google-ads/answer/6167122?hl=en-ZA&ref_topic=10287125&visit_id=… support.google.com
Share the Post:

Related Posts