RSA vs. SKAG Google Ads: Which Structure Delivers Lower CPA for Local Campaigns?

The Real Question Behind the Structure Debate

Every local service advertiser eventually faces this fork: build tight Single Keyword Ad Groups (SKAGs) or Single Theme Ad Groups (STAGs) for control, or lean into Google's Responsive Search Ad (RSA) machine-learning engine with broader keyword inputs?

The debate sounds tactical. It isn't. It's a budget-leak question. The wrong structure for your campaign's current data maturity silently siphons spend into irrelevant match expansions — clicks from people who were never going to book your service. For a local plumber, HVAC company, or law firm, those wasted clicks add up fast.

This teardown walks through both structures, what the data actually tells us, and a decision framework so you can pick the right setup — or the right mix — for where you are today.

How Each Structure Works (and Where Each Breaks)

RSA with broad-input theming loads 8–15 headlines and 4 descriptions into a single ad, then lets Google's algorithm test combinations and expand match coverage. The intended win: machine learning finds converting queries you'd never have thought to bid on. The real risk for local advertisers: without tight negative keyword lists and at least several weeks of solid conversion data, Google's matching logic reaches far outside your geography and service scope.

SKAGs / STAGs do the opposite. One keyword (or a tightly themed cluster) per ad group, with ad copy that mirrors that keyword almost verbatim. This forces a near-perfect relevance triangle — keyword → ad copy → landing page — which is exactly what Quality Score rewards.

Where SKAGs break: they require more build time, more ongoing maintenance, and they starve Google's algorithm of the cross-keyword signal volume it needs to optimize bidding. In small local markets with limited monthly search volume, a pure SKAG build can fragment your data so thin that Smart Bidding never learns properly.

Quality Score, CTR, and the CPA Math

Quality Score (QS) is the lever that ties these structures to your actual cost. Google publicly confirms that QS affects both Ad Rank and the effective CPC you pay — a higher QS means you pay less per click for the same position.

A useful way to think about this: for every 1-point improvement in Quality Score, you can expect a meaningful reduction in cost-per-click (Google's own documentation confirms QS impacts auction pricing, though exact per-point discounts vary by auction). As a rough illustrative model, moving a keyword from QS 4 to QS 7 can shrink your effective CPC by 20–35% — which flows directly into CPA.

Tightly themed ad groups (SKAGs/STAGs) tend to produce higher QS because the keyword-to-ad copy relevance is explicit. Broad RSA inputs, particularly when paired with broad match, can dilute that relevance across too many query variants, pulling QS down on individual keywords.

The CTR dimension: RSAs can sometimes win on CTR in competitive auctions because Google serves the most contextually relevant headline combination per user. But CTR without conversion relevance is a vanity metric. A local HVAC company that sees strong CTR on queries like 'how to fix AC yourself' is just funding Google's revenue, not their own.

The CPA math that actually matters:

• Higher QS → lower CPC → same budget buys more clicks
• Tighter match → higher conversion rate from clicks that do arrive
• Both factors compound: a 20% CPC reduction plus a 15% conversion rate lift can cut CPA by roughly 30% in a simple model (illustrative, based on the compound formula: new CPA = old CPA × 0.80 × 0.85 ≈ 0.68× old CPA)

Where RSAs Leak Budget for Local Advertisers

The most common budget-leak pattern we see in local RSA campaigns:

1. Broad match expansion into non-local queries. A roofing company in Austin starts matching 'roofing companies' nationally. Impressions balloon, CTR craters, QS falls. 2. Headline combination mismatch. RSA serves an unexpected headline combo that doesn't match the landing page's core message — QS drops on Expected Landing Page Experience. 3. Thin negative keyword lists. RSA's reach requires aggressive negatives. Without them, service queries you can't fulfill eat your daily budget by 9 AM. 4. Premature Smart Bidding on low-volume data. Target CPA bidding needs roughly 30–50 conversions per month to stabilize (Google's own guidance suggests 50 conversions/month as a threshold for reliable Smart Bidding performance). Local campaigns in smaller markets often can't hit that, so the algorithm oscillates wildly.

This is why we recommend reading our piece 'Ad Consolidation vs Segmentation: Lower CPA for Local Ads' alongside this one — the consolidation question and the RSA-vs-SKAG question are two sides of the same decision.

The Decision Framework: Which Structure for Your Campaign?

Use this three-question filter before choosing your structure:

1. How many monthly conversions are you currently tracking?

• Fewer than 20/month: lean STAG (tightly themed, 3–5 keywords per group), exact and phrase match dominant, manual CPC or Max Clicks. Protect data concentration.
• 20–50/month: hybrid — consolidate to STAGs, layer RSA with pinned headlines for brand/service terms, introduce Target CPA cautiously.
• 50+/month: RSA-forward structure with consolidated ad groups and Smart Bidding can now learn reliably.

2. How competitive is your local market? High-competition local markets (personal injury law, emergency plumbing, HVAC) have expensive CPCs — every Quality Score point is worth more in dollar terms. In those markets, STAG structure with pinned RSA headlines (pin Headline 1 to your service + city) preserves relevance without abandoning RSA format requirements.

3. How mature is your negative keyword list? If you're early in a campaign with fewer than 100 negatives built out, broader RSA + broad match is a budget fire. Build your negatives first. Then expand match type coverage.

For context on how your ad structure decision connects to full-funnel efficiency, see our articles 'LTV:CAC Ratio Benchmarks for Local Service Businesses' and 'Second Ad Channel: What It Does to Your Blended CAC' — structure choices here ripple into your blended acquisition cost across every channel.

The Pinned RSA: The Practical Middle Ground

The approach that threads the needle for most local advertisers isn't pure SKAG or pure broad RSA — it's the pinned RSA inside a tightly themed ad group.

Here's what that looks like in practice:

• Ad group theme: 'Emergency Plumber [City]'
• Keywords: exact and phrase variants of that theme only
• RSA setup: Pin Headline 1 = 'Emergency Plumber in [City]' (locks relevance), leave Headlines 2–3 and all descriptions unpinned for Google to test
• Negative keywords: 30+ negatives before launch, reviewed weekly for the first 60 days

This structure preserves the QS benefits of explicit keyword-to-ad relevance (because Headline 1 always matches), while letting Google's algorithm optimize the supporting copy. You're not handing the machine a blank canvas — you're giving it guardrails.

As a labeled illustrative model: if this approach moves a campaign from QS 5 average to QS 7 average on its core keywords, and your current CPC is $18, that shift could reasonably lower effective CPC to the $13–$15 range — translating to 15–25% more clicks from the same daily budget, with no change in spend.

Bottom Line: Structure Is a Budget Decision, Not a Preference

The RSA-vs-SKAG debate only has a correct answer relative to your campaign's data maturity, market competition, and negative keyword discipline. A broad RSA structure in an immature local campaign is one of the fastest ways to burn budget with nothing to show for it. A pure SKAG build in a data-starved market fragments your signal and blocks Smart Bidding from ever learning.

The practical hierarchy:

• Control structure first (STAGs, strong negatives, exact/phrase match)
• Layer RSA with pinned headlines to preserve QS
• Earn your way into broader match expansion as conversion volume grows
• Audit match type reports weekly for the first 90 days — that search term report is your budget-leak detector

Want a second set of eyes on your current campaign structure? Nika Spark runs a paid-ads diagnostic specifically for local service businesses — we map your account's QS distribution, match-type exposure, and CPA by ad group, then show you exactly where the leaks are. Book a call to see if it makes sense for your account.