Why Multi-Location Glaucoma Groups Run Broken Structures

The structural choice is rarely deliberate. Most multi-location glaucoma groups inherit a campaign architecture that grew organically over a decade, with a campaign per location added each time a new office opened and never consolidated when volume turned out lower than expected. The result by 2026 is an account with 8 or 12 location campaigns averaging 5 to 12 conversions each per month, none of which clear the threshold tCPA needs to optimize. Smart Bidding sits in a permanent learning state, and the practice’s marketing lead reports CPA volatility that the agency attributes to seasonality.

The opposite mistake is just as common in the segment. A multi-location group consolidates everything into a single campaign with location bid adjustments, which lets Smart Bidding work but loses the ability to ad-copy or landing-page differentiate by location. The patient sees a generic ad that does not reference the closest office and lands on a hub page that requires a click-through to find the right address. Conversion rate drops 15 to 25 percent against a properly differentiated structure, even though the bid algorithm is technically functional. The right structure is rarely either extreme; it is the one that matches the data density at each location to what Smart Bidding needs.

How the 30-Conversion Threshold Decides the Structure

The decision rule is simple in principle. Smart Bidding tCPA needs roughly 30 conversions per month per campaign to optimize stably; tROAS needs 50. Below those thresholds, the bid strategy spends a disproportionate share of its time in learning mode and produces the volatility that gets blamed on seasonality. The right structure follows the data density at each location, not the organizational chart of the practice.

Practical application for a glaucoma group looks like this. Locations under 30 conversions per month roll into a consolidated campaign with location bid adjustments and location-specific ad customizers (the city name in the headline, the closest office address in the description). Locations between 30 and 80 conversions per month get their own campaigns, but those campaigns share a portfolio bid strategy so the algorithm can pool signal across the portfolio while still letting each campaign carry its own creative. Locations above 80 conversions per month get fully independent campaigns with their own bid strategies, because the data density is high enough to support standalone optimization.

The discipline is what separates accounts that work from accounts that look organized but underperform. The structural rule has to be documented and applied consistently, including when a new location opens. New locations almost never have 30 conversions per month in the first 3 to 6 months; launching them as standalone campaigns from day one creates the very fragmentation the rule exists to prevent. The companion PPC benchmarks for eye care piece covers the volume thresholds in more detail.

The Threshold Numbers and a Documented Multi-Location Case

Smart Bidding thresholds: tCPA needs 30 or more conversions per month; tROAS needs 50 or more (Yeezypay, January 2026). These are floors, not targets, and accounts that hover just above the floor experience more volatility than accounts that clear it comfortably. For a glaucoma group the implication is that each campaign needs to be sized to durably exceed 30 monthly conversions across season variation, not the average of an annualized count.

The documented multi-location case in eye care comes from Pacific Cataract & Laser Institute, 17 locations across 6 states, running a consolidated structure with location bid adjustments and ad customizers. The reported result was an $82 CPA below a $100 cap with 44 percent impression share (Cardinal Digital Marketing, February 2025). The relevant detail is not the absolute numbers, which depend on the practice’s vertical and competitive density, but the structural choice. A 17-location group running consolidated rather than per-location is the opposite of what most multi-location glaucoma groups have built, and the result is meaningfully better. The companion portfolio bidding piece covers the bridge structure for the mid-volume case.

Five Red Flags in Multi-Location Glaucoma Account Structure

The following patterns indicate the account is built on assumptions that no longer match Smart Bidding requirements. First, every location has its own campaign regardless of volume. The agency built one campaign per office because that was the conventional structure circa 2018 to 2021, before tCPA’s data-density requirements tightened. The structure looks tidy in the campaign list and produces 8 or 12 campaigns with chronically insufficient conversion volume each. Second, no portfolio bid strategy across low-volume campaigns. Even when the campaigns exist, they bid independently, which means signal that could be pooled is wasted.

Third, location keyword cannibalization between campaigns. Two location campaigns bid against each other on a generic glaucoma query because the targeting overlaps and neither campaign has been excluded from the other’s auctions. The practice pays a CPC premium for impressions it would have won at lower cost without the internal competition. Fourth, brand campaigns are not separated from non-brand at the location level. A user searching for a specific surgeon by name triggers the location’s non-brand campaign at full CPC, which is both expensive and a signal-quality issue for the bid algorithm.

Fifth, no structural review when a new location opens. The practice opens a third or fourth office, the agency adds a campaign by reflex, and nobody asks whether the new location can sustain 30 monthly conversions or whether the campaign should join the consolidated rollup until volume builds. Six months later the new campaign has 8 conversions per month and Smart Bidding is in permanent learning mode for that location. The companion PPC agency red flags piece covers the structural-question discipline that should be in any agency QBR.

The Concrete Audit and Restructure Sequence

Pull the trailing 90 days of conversion volume by location campaign. Annualize and adjust for seasonality. Any location averaging under 30 conversions per month durable across the seasonal range gets merged into a consolidated campaign. The merge is operationally painful (history is preserved through the new campaign’s lookback, but learning starts over) and the practice’s reporting has to bridge the structural change cleanly. Most agencies underestimate the reporting cleanup work, which is why structural restructures get deferred.

Mid-volume locations (30 to 80 conversions per month) get their own campaigns but share a portfolio bid strategy. The portfolio lets the algorithm pool signal across the portfolio while each campaign retains its own creative, ad customizers, and landing page. High-volume locations (above 80) stand alone with independent bid strategies, because the data density supports it.

Document the rule. Every new location follows the rule. New locations launch inside the consolidated campaign and earn promotion to a dedicated campaign only after they sustain 30 conversions per month for two consecutive months. The rule prevents the slow accumulation of low-volume campaigns that turns multi-location accounts into structural messes over a 36-month horizon. The companion 15-minute Google Ads audit piece covers the quick-look version of the volume audit.

Specialty Vision’s Take on Multi-Location Account Structure

Our view, account structure is the closest thing PPC has to an infrastructure decision. Get it wrong and every subsequent optimization is a workaround layered on top of the wrong base; get it right and Smart Bidding does most of the work the agency would otherwise spend hours on each week. Most multi-location glaucoma groups are running structures built for 2022 data densities, when tCPA was more forgiving and Smart Bidding learning periods were less sensitive to thin-volume campaigns.

The harder problem is that restructure work is invisible to the practice’s marketing lead. The reported metrics may not move for 30 to 60 days post-cutover because Smart Bidding has to relearn against the new structure, and the cleanup looks like effort with no immediate result. Agencies avoid recommending it because the conversation is uncomfortable, and practice leads avoid asking because the topic is technical.

The structural review should happen annually as a forced ritual, not when a problem becomes acute. For the broader audit context, see our 2026 PPC audit playbook.