Why Self-Cannibalization Is Common in Multi-Location Accounts

In multi-location eye care groups, the practice’s own locations are often bidding against each other in the same auction. The operator is raising their own CPCs while Google quietly takes the margin. This piece shows how to find and fix internal cannibalization, and why most multi-location agencies never look for it.

The structural reason is that location-level campaigns get built independently as new offices open or are acquired. Each location’s campaign uses similar keyword lists because the services are similar, and the geo targeting tends to overlap because catchment areas overlap in urban and suburban markets. The result is two or more campaigns competing for the same searcher in the same moment, with the practice paying both bid prices through the auction. Without an explicit audit at the account or MCC level, the pattern stays invisible because individual campaign reporting looks healthy on its own. Each campaign hits its targets; the aggregate pays a tax to itself that nobody notices.

How Cannibalization Shows Up in the Account

Cannibalization happens when two or more campaigns from the same account, or the same MCC, bid on overlapping keywords with overlapping geo targeting. Google rarely flags it explicitly. The symptoms show up as inflated CPCs, fragmented Quality Score, and auction-insights reports that list the operator’s own other campaigns as top competitors. The signal is there, just not in the dashboard the operator usually looks at.

Auction Insights is the diagnostic tool. At the campaign or account level, the report shows which competitors appear in the same auctions as the campaign. When the operator’s own domain or related campaigns appear in the report, that is direct evidence of internal competition. The pattern is most obvious on non-brand keywords with broad geo targeting. Brand campaigns are sometimes affected too, particularly when “[brand] [city]” variants are bid by multiple locations, each trying to rank for searchers in nearby cities they consider part of their catchment.

The cost is paid through CPC inflation rather than visible waste. The operator does not see a “cannibalization spend” line item; they see CPCs that drift upward over months, attributed to “competitive pressure” or “market dynamics,” when the real driver is internal. The fix changes the cost structure permanently because the auction price drops once the internal competition is removed. For broader audit context, see 15-minute Google Ads audit.

Typical Impact and Account-Size Patterns

Typical impact on multi-location accounts is 10 to 25 percent CPC inflation on cannibalized keywords. The exact figure depends on bid density, geo overlap percentage, and how many locations participate in the cannibalization. Two-location groups with mild geo overlap see modest inflation; five-or-more-location groups in dense urban markets often see the upper end of the range, sometimes higher.

Accounts with five or more locations running separate per-location campaigns tend to show the highest cannibalization rates. The structure is intuitive (each location gets its own campaign for control and reporting) but the auction mechanics punish the structure when geo targeting overlaps. Urban markets where multiple locations share commute-distance catchment are the worst-affected because the geo overlap is the largest. Rural and suburban networks with non-overlapping catchment areas often see no cannibalization, which is why the problem is concentrated in metro multi-location groups. The fix value scales with the number of overlapping campaigns and the keyword density. The patterns that produce the highest cannibalization are urban markets where ZIP catchments overlap by 30 percent or more, branded campaigns where sister locations bid on each other’s brand-plus-city queries, and service-line campaigns that share keyword sets across locations without any geo separation. Cannibalization compounds without an audit because each new location added to the account inherits the same structural overlap, and the typical timeline from initial setup to material CPC inflation runs 12 to 18 months before anyone notices the trend in reporting. For account structure context, see account structure for multi-location.

Five Red Flags in Multi-Location Account Setup

The following patterns indicate cannibalization is likely present and worth auditing. Each one is a specific signal an operator can check in 15 to 30 minutes.

First, auction insights shows the operator’s own account as a top-5 competitor. The single clearest signal of self-bidding. If the practice’s domain appears in its own auction insights report at meaningful impression share overlap, internal competition is real. Second, two or more locations bid on identical non-brand keywords with overlapping geo. The structural setup that produces cannibalization. Third, brand campaigns from different locations bid on “[brand] [nearby city]” variants. Each location trying to capture branded searchers in adjacent markets ends up competing on those queries.

Fourth, CPCs on branded terms trending up despite no external competition change. Branded CPC inflation without a new external competitor is almost always internal. Fifth, no account-level cannibalization audit ever run. The most common state, and the reason cannibalization often goes uncorrected for years. The audit is straightforward and the fix is structural, but if nobody is tasked with running it, it does not get run. Sixth, search-term reports across multiple campaigns showing identical converting queries, which means the same searcher is producing conversion credit in two campaigns at once and the operator is paying twice for the same auction-winning click. Seventh, MCC-level spend patterns where branded budget grows materially faster than non-brand without a strategic rationale, since the most common driver of branded inflation is internal bidding. For audit cadence context, see weekly search term reports.

The Audit and the Structural Fixes

Run the audit at the account level. Pull auction insights and filter to any “self” matches where the operator’s own domain or related campaigns appear. Map which campaigns are cannibalizing which, and on which keyword sets. The map is the fix-design document; without it, the operator is guessing at structural changes that may or may not address the actual overlap.

Fix options vary by structure. Consolidation into a single campaign with location bid adjustments is the cleanest solution for groups in dense markets, since one campaign cannot bid against itself. Tightening geo targeting so locations do not overlap works when catchment areas can be cleanly separated by ZIP boundaries. Adding location-specific negatives (excluding a neighboring location’s city name from the competing campaign) is the lightest-touch fix and works for surgical problems where most of the overlap is on a small set of geo terms. Most multi-location accounts need a combination of all three. Negative keyword discipline catches what structural fixes miss; see quarterly negative keyword audits for the cadence.

Execution starts with exporting auction insights at the account level monthly so the trend line is visible. The team builds a location-overlap matrix listing each location’s geo targeting, top non-brand keyword set, and brand variants, then highlights cells where locations share both geo and keyword. Staged consolidation starts with overlapping non-brand terms because those produce the largest CPC drop, then proceeds to brand-plus-nearby-city variants.

Specialty Vision’s Take on Cannibalization Audits

Our view, cannibalization is invisible unless the operator explicitly looks for it, and most multi-location agencies do not look. The metric never appears in standard reporting, the platform does not flag it, and the agency has no commercial incentive to surface a problem they helped create. An annual cannibalization audit typically recovers 10 to 20 percent of spend that was funding internal competition. The fix is structural and pays back forever once implemented because the auction price drops permanently when the internal competitor is removed. We recommend annual cadence at minimum, semi-annual for fast-growing groups. The work happens once, the savings persist, and the operator should expect the audit to pay for itself many times over within the first quarter. For broader audit framework, see our 2026 PPC audit playbook.