Google Ads Location Targeting: How Senior Media Buyers Prevent Wasted Budget Under Google’s AI Automation

Google Ads Geotargeting is no longer just a checkbox in your campaign setup. Under Google’s modern AI-driven ecosystem, default location settings have become one of the quietest yet most destructive ways budgets leak.

Google Search is no longer a simple keyword-matching engine; it is a semantic understanding and user-experience platform. When automated bidding (Smart Bidding) is paired with default location settings, the algorithm prioritizes driving the cheapest conversions on paper—even if those conversions originate from low-value regions with zero lifetime value (LTV).

This guide bypasses UI basics like "how to click the pencil icon" and focuses on the technical realities, algorithmic behaviors, and structural adjustments required to maintain absolute control over your geographic budget.

The Modern Geotargeting Landscape: Privacy Sandboxes, IP Protection & Masked Locations

To build an airtight geotargeting strategy, you must first understand the technical signals Google uses to determine user location and how those signals are degrading.

Traditionally, Google has relied on IP addresses, GPS data, device settings, and search queries containing location terms to pinpoint users. However, the rise of strict privacy protocols has significantly compromised IP-based targeting:

• iCloud Private Relay: Apple automatically masks the IP addresses of Safari users on iOS and macOS, routing traffic through two separate relays. This masks the user's exact location, often reverting it to a broad, regional level (e.g., state-level instead of city-level).

• Chrome IP Protection (Gnatcatcher): Google’s own privacy initiatives mask user IP addresses to prevent cross-site tracking.

• VPN Ubiquity: A growing percentage of high-value, tech-literate, or B2B audiences run continuous VPNs, entirely spoofing their physical locations.

The Impact on Your Campaigns

When Google cannot accurately resolve a user’s physical IP, it bucket-assigns them to a "Masked" or "Unknown" location state.

If your campaigns rely heavily on hyper-local parameters (such as narrow radius targets), these privacy-masked users are either completely excluded from your funnel (increasing your CPMs due to artificially shrunk audience sizes) or misclassified into broader zones, causing immediate budget leakage.

To combat privacy-driven data loss, enterprise buyers must layer geographic constraints with a robust Google Ads audience strategy and precise Google Ads demographic targeting to maintain high intent.

Stopping the "Location Leak": Navigating Presence vs. Interest in the Era of AI Max

The single most common setting mistake in enterprise Google Ads accounts is leaving the Target setting on its default: Presence or Interest (Recommended).

To help you audit your setup, here is how the two primary settings compare under a media buyer's lens:

1. Default Setting: Presence or Interest (Recommended by Google)

• Mechanism: Targets people in, regularly in, or who have shown interest in, your targeted locations.

• Best Suited For: Travel, hospitality, or global brands targeting tourists/out-of-town buyers.

• The Performance Trap: Highly prone to major budget waste. For example, an out-of-market user in another country searching for "commercial litigation attorney New York" will trigger your high-CPC local ad.

2. Advanced Setting: Presence (Highly Recommended for Lead Gen/Local B2B)

• Mechanism: Targets strictly people physically located in or regularly present in your targeted locations.

• Best Suited For: Local lead generation, physical service providers, retail stores, and regional B2B brands.

• The Protection: Acts as a hard geographic boundary, restricting ad delivery exclusively to verified physical users within your target area.

The Performance Max & AI Max "Leak"

In automated campaign types like Performance Max and modern AI Max variations, Google’s machine learning is designed to find conversions at all costs. If you select Presence or Interest, the algorithm will actively bypass your geographic targets under the guise of "search interest." It targets users hundreds of miles away simply because they looked up a localized term, rapidly inflating your CPA without delivering actual business value.

Actionable Defense Protocols

1. Switch to "Presence": Go to your campaign settings and explicitly select Presence: People in or regularly in your targeted locations.

2. Deploy Account-Level Exclusion Lists: Do not rely solely on target settings. Upload a comprehensive exclusion list of countries, states, or high-cost/low-yield ZIP codes at the Account Level. Account-level exclusions act as a hard boundary that even Performance Max and AI Max algorithms cannot cross.

Probabilistic Location Mapping vs. Final Revenue

Under the hood, Google does not always rely on hard, deterministic signals (like GPS) to match locations. Instead, it utilizes Probabilistic Location Mapping. When privacy walls degrade user tracking signals, Google’s machine learning fills the gaps by "predicting" user location based on historical search patterns and peripheral intent.

Here lies the core conflict of interest for senior media buyers:

• The Algorithm's Goal: Optimize for Proxy Conversions (e.g., cheap form-fills, low-cost leads) to hit your automated tCPA target on paper.

• The Buyer's Goal: Optimize for Final Business Revenue (LTV).

Because "Presence or Interest" allows probabilistic matching, Smart Bidding will systematically route your budget to out-of-market users who exhibit search behavior resembling your target profile, but who reside in non-serviceable areas. They convert cheaply on your landing page, inflating your conversion rates on the Google Ads dashboard, but land in your CRM as 100% dead, unserviceable leads. Switching to strict "Presence" is the only baseline mechanism to break this algorithmic feedback loop.

Advanced Strategy: Tiered Geotargeting (Geo-Splitting) Under Smart Bidding

When using Google Smart Bidding strategies like Target CPA (tCPA) or Target ROAS (tROAS), treating a large geographic territory (e.g., the entire United States) as a single target creates a systemic bidding trap.

The Smart Bidding Trap

Smart Bidding looks for the easiest path to hit your CPA goal. If you group high-value, high-competition regions (e.g., New York, California) in the same campaign as lower-cost, lower-value regions (e.g., Ohio, Kentucky), the algorithm will naturally shift your budget toward the cheaper regions because they yield cheaper conversions.

However, for most businesses, a lead from Manhattan is worth 5x more than a lead from rural Ohio. By letting the algorithm optimize for a flat, national CPA, you are systematically starving your most profitable markets of budget.

The Illusion of Location Bid Adjustments in Smart Bidding

A critical mistake many experienced buyers make when attempting to optimize geographic performance is relying on Location Bid Adjustments (e.g., setting a -50% adjustment on an underperforming state) while running automated bidding (tCPA/tROAS).

Here is the technical reality: Under Smart Bidding, location bid adjustments do not lower your actual CPC. Instead, Google’s algorithm interprets a negative bid adjustment as a signal to modify your target CPA/ROAS for that specific region.

• The Trap: If your account-wide tCPA is $100, and you apply a -50% bid adjustment to state X, you are telling the algorithm: "I want leads from State X at a $50 CPA."

• The Result: Instead of reducing your ad delivery in State X, the algorithm lowers its quality threshold. It begins targeting the lowest-quality, cheapest, and most spam-heavy traffic in that region to force the conversion cost down to $50 artificially.

To control geographic distribution under Smart Bidding, you cannot rely on bid sliders. You must enforce hard budget constraints. This requires a structural transition to Geo-Splitting; physically separating high-LTV regions from low-LTV regions into dedicated campaigns where budgets are hard-capped.

The Solution: The "Geo-Splitting" Framework

Instead of running a single national campaign, segment your budget using a tiered approach. This architectural layout controls how budgets flow between high-value and low-value regions:

• Tier 1 (Premium Markets): Higher customer lifetime value (LTV). Allocate dedicated budget here and set a higher, more aggressive tCPA or lower tROAS target to win dominant impression share in these highly competitive zones.

• Tier 2 (Average Markets): Moderate competition. Set standard CPA/ROAS targets.

• Tier 3 (Budget/Low-Yield Markets): Low CPCs, but low conversion values. Run this on a strict, limited budget with a highly restrictive tCPA to capture cheap volume without draining your core budget.

This tiered structural allocation is a mandatory protocol for scaling Google Ads ecommerce campaigns where product margins vary wildly by state.

Industry Performance Benchmarks

According to multi-brand account audits, splitting campaigns into distinct geographic tiers yields an average 18% to 22% increase in actual bottom-line ROI. While the dashboard CPA might appear slightly higher due to increased competition in Tier 1, the overall lead-to-revenue conversion rate drastically improves.

Technical Update: Transitioning to Asset-Based Location Groups

For multi-location brands, franchise networks, or local service businesses, handling locations manually is incredibly inefficient. Google has deprecated legacy, feed-based location groups and transitioned fully to Asset-Based Location Groups.

Here is how the legacy layout compares to the updated infrastructure:

How to Leverage Asset-Based Location Groups

1. Link via Account Assets: Ensure your Google Business Profile (GBP) is linked directly to your Google Ads account via the Assets tab rather than a legacy feed.

2. Create Location Groups using Store Codes: Categorize your physical locations using specific labels or store codes. For example, group all "Metropolitan Express" locations under one asset group and "Suburban Hubs" under another.

3. Deploy in Smart Campaigns: You can now instantly attach these Asset-Based Location Groups to PMax or Local campaigns, ensuring your local assets (address, local phone number, and directions) dynamically swap based on proximity to the user.

Decoding the "Unknown" Location Group in Matched Reports

As tracking protections tighten, your location reports will show a growing segment labeled "Unknown". This represents users who matched your keyword criteria but whose exact location signals were masked.

Because you cannot directly target or exclude "Unknown" locations in the settings panel, you must follow a systematic routing process to evaluate their value:

If the "Unknown" segment is draining budget with zero returns, it is usually a sign that your match types are too broad. Because Google cannot verify the physical location, it relies heavily on search intent; if your keywords are loose, you are likely bidding on out-of-market searchers.

Advanced Diagnostic: Is It Privacy-Driven Data Loss or Click Fraud?

When auditing the "Unknown" geographic segment, senior buyers must differentiate between two distinct types of untracked traffic:

1. Type 1: Privacy-Driven Unknown (High-Value): High-intent B2B or premium consumers using corporate VPNs, iCloud Private Relay, or strict privacy browsers. This traffic typically exhibits high Engagement Rates and solid Session Durations in GA4.

2. Type 2: Fraud-Driven Unknown (Ad Waste): Click farms, scrapers, or bot networks operating through masked proxies, often pushed via the Google Search Partners network. This traffic is characterized by a near-100% bounce rate and sub-second session durations.

The Action Plan for Media Buyers

Do not make the mistake of ignoring "Unknown" performance. Build a custom report in Looker Studio blending Google Ads location data with GA4 engagement metrics:

• If the "Unknown" segment shows healthy engagement: Maintain distribution. These are highly guarded, high-value human prospects.

• If the "Unknown" segment shows low engagement (<15%): Do not attempt to exclude the "Unknown" location directly (Google's UI does not support this). Instead, immediately opt out of Google Search Partners at the campaign settings level, shift your keywords from Broad Match to tight Phrase/Exact match, and aggressively build out your Google negative keyword lists. This starves the bot networks of the loose semantic matches they rely on to trigger your ads.

Hyper-Local Personalization: Dynamic Location Insertion (DLI) & Split-Testing

High CTRs tell the algorithm your ad is highly relevant, which lowers your CPC via an improved Quality Score. One of the easiest ways to scale CTR is through Dynamic Location Insertion (DLI).

Using location insertion in your Responsive Search Ads (RSAs) allows Google to dynamically inject the user's city, state, or country directly into your ad copy.

Implementation Best Practice

Instead of writing static copy like:

"Affordable Commercial Roofing Services"

Use dynamic syntax:

"Commercial Roofing in {Location.City:New York}"

If Google cannot resolve the user's city (due to the privacy protocols outlined in Section 1), it will automatically fall back to your default text ("New York"). If it resolves the user's location as "Brooklyn," the ad dynamically reads: "Commercial Roofing in Brooklyn."

Scientific Split-Testing by Region

Do not assume every geographic market responds to the same value proposition. Use campaign drafts and experiments to split-test creatives:

• Urban/Metro Areas: Test messaging centered around speed, convenience, and 24/7 availability (e.g., "Same Day Service").

• Suburban/Rural Areas: Test messaging focused on trust, community reputation, and family-owned values (e.g., "Serving Our Community Since 1998").

Conclusion

In the era of automated bidding and privacy-first data loss, location targeting is no longer a "set-and-forget" configuration. Treat it as an active variable that requires continuous architectural oversight.

Leaving your campaigns on Google’s default geotargeting settings is an open invitation for Smart Bidding algorithms to prioritize cheap, low-value out-of-market conversions over high-LTV localized revenue. By shifting your accounts to strict "Presence" targeting, structurally segmenting your geographic territories into value-based tiers, and actively auditing the "Unknown" data segment, you reclaim control of your media spend.

Stop letting automated algorithms dictate where your budget goes. Audit your geographic parameters, deploy hard account-level exclusions, and ensure every dollar is routed to territories that drive real bottom-line business growth