The buried setting that costs the most

Multi-location eye care groups consistently make the same four geo-targeting mistakes, and each one costs 10 to 20 percent in wasted spend. The patterns are well-documented but almost never audited because geo settings live three clicks deep in the campaign UI, behind defaults that look fine on the surface and quietly bleed budget underneath. A cataract practice running $30,000 a month across five locations and losing 15 percent to geo waste is leaving $54,000 a year on the table without any visible signal in the standard dashboard.

The waste does not show up as a campaign that obviously fails. It shows up as efficient-looking campaigns that are spending part of their budget in markets the practice does not actually serve, or competing against the practice’s own other locations, or paying premium urban CPCs to reach searchers who would never travel to the clinic. The diagnostic is straightforward, but it requires opening the geo settings on every campaign and checking them against actual patient origin data from the EMR.

The two layers of geo targeting most accounts confuse

Google Ads geo targeting has two layers that operate independently. The first is location targeting, which controls who can see the ad based on where they are. The second is location bid adjustments, which controls how aggressively you bid by location for the audiences you have already targeted. Most accounts have one layer set up reasonably and the other left at default, which is rarely the right combination.

Location targeting is the bigger lever. Most accounts misuse it in one of two directions. Either they set radius targeting that spills into competitor territory and into ZIPs the practice has no real draw from, or they blanket the entire state or DMA without acknowledging that the practice serves only a subset of it. Both errors waste spend, but they waste it differently. Radius spillover wastes spend on auctions where the practice is not the right answer; blanket targeting wastes spend on impressions that will never become patients.

Location bid adjustments are the smaller but compounding lever. Even with correct location targeting, treating every ZIP equally leaves money on the table. Some ZIPs convert at 2x the practice average; others at half. Bidding equally across them subsidizes the weak ZIPs with the strong ZIPs’ efficient performance. The fix is per-ZIP bid adjustments based on EMR-validated patient origin data, which most agencies skip because it takes 4 to 6 hours per location per quarter to do well.

The waste estimate and why CPC volatility matters

No public benchmark exists on geo-waste specifically, but practitioner data suggests 10 to 25 percent of multi-location PPC spend in broadly-targeted campaigns lands outside the practice’s actual service area. The Patient10x August 2025 data on urban-versus-rural CPC differentials (urban CPCs running 2-3x rural) compounds the problem; geo precision affects efficient spend beyond just relevance. Paying $60 for a click in the wrong urban ZIP wastes more than paying $20 for an equivalent wasted click in a rural one.

The waste scales with account size. A $1M portfolio with 15 percent geo waste leaks $150,000 a year that could be redirected to the right ZIPs at the right bids. Urban cataract patients typically travel 5 to 10 miles for a consultation, suburban patients 10 to 25 miles, and rural patients 30 to 60 miles. EMR ZIP analysis differs from Google Analytics geo data in important ways: the EMR captures the address the patient gave at intake, while GA reports IP-derived device geography that can drift dozens of miles for mobile and VPN traffic. PE-backed groups operating multiple urban locations consistently see geo overlap in the 15 to 25 percent range when measured against EMR ZIP distribution, with two locations paying separately for the same searcher in a shared catchment ring. The multi-location account structure guide and 15-minute Google Ads audit usually surface the problem within the first hour. The fix typically pays back within two months.

The four common mistakes worth checking

The first common mistake is radius targeting larger than the actual catchment. The default 20 or 25 mile radius assumes patients will travel that far, which they often will not for routine cataract or refractive consultations. Most practices’ actual patient origin clusters within 8 to 12 miles of the clinic, with longer-distance patients arriving via referral rather than ad click. Setting the radius to the actual EMR-validated range cuts impression waste immediately.

The second is the “people interested in” target type, which is enabled by default in many account configurations. This setting matches the ad to people researching the geographic area, not to people actually in or regularly in it. A LASIK searcher in another state researching surgeons in your city sees your ad. They are not your patient. The fix is switching to “People in or regularly in your targeted locations” – a single dropdown change with significant spend impact.

The third is overlap between locations with no exclusion rules. The Phoenix campaign and the Scottsdale campaign both target their respective metros with overlapping radii; they bid against each other in the overlap zone, inflating CPCs in auctions the practice would have won at lower cost with a single coordinated bid. The fourth is no bid adjustments for high-value versus low-value ZIPs, leaving the spend distributed equally across ZIPs that perform unequally. The fifth (a corollary) is trusting the mobile geo signal without validation; mobile location data is noisier than desktop and worth a separate audit. The PE-backed governance piece covers the portfolio-level approach to all of this.

The audit and the fixes that compound

Audit each campaign’s geo settings. Set the target type to “People in or regularly in your targeted locations” – the default is usually “Interested in,” which is rarely what an eye care practice actually wants. Pull EMR data on actual patient ZIPs over the last 12 months and match the targeted radius to the data, not to the assumption. If 90 percent of your patients come from within 10 miles, the 25-mile radius is wasting money on the outer 15 miles.

Pull patient ZIP distribution directly from the EMR rather than approximating from third-party tools. In Epic, the report lives under the registration data extract or practice analytics module. In Modernizing Medicine (EMA), the patient demographics export by date range gives the same picture in CSV. Nextech offers an analogous export. Each pull takes 15 to 30 minutes and produces the ZIP-by-volume picture bid adjustments should be built against. Add negative location exclusions for neighboring locations’ primary ZIPs to prevent internal cannibalization. The Phoenix campaign explicitly excludes Scottsdale primary ZIPs and vice versa. Bid-adjust the top decile of ZIPs upward by 20 to 30 percent based on EMR-validated conversion data. Reduce the bottom quartile by 10 to 20 percent, and exclude the lowest performers entirely when 90-day conversion volume stays at zero. The audit takes 2 to 4 hours per location per quarter; savings typically run 5 to 10x the time investment.

Specialty Vision’s take on geo as the lazy management tell

Our view is that geo targeting is where lazy account management hides most reliably. Anyone can set a 25-mile radius and call the campaign launched. Doing geo correctly requires patient-data analysis from the EMR and location-level bid management, work that runs 4 to 6 hours per quarter per location and that most agencies skip because nobody asks for it. The savings typically pay for the work three to five times over within the first quarter, and the improvements compound as patient origin patterns shift over time. We treat the geo audit as the highest-ROI hour an auditor can spend on a multi-location account. For the full audit framework that this fits inside, read The 2026 PPC Audit Playbook for Specialty Eye Care Practices.