A country appears in an app acquisition report that the growth team did not intend to buy. One person proposes changing a location setting; another wants separate campaigns; a third calls the traffic fraudulent. Before approving any of those responses, establish what the country column actually measures and which campaign produced the activity.
Google App campaigns location targeting should be audited against three separate questions: which locations the campaign is configured to reach, where each reporting system places the observed activity, and which customers the business can serve profitably. A mismatch between those answers requires investigation. It does not, by itself, prove a targeting failure, invalid traffic or a need to rebuild the campaign.
This guide provides an editorial decision framework for established app teams choosing whether to correct settings, repair reporting, separate markets or escalate a delivery concern. It is not an account-specific diagnosis or a promise of geographically exact delivery.
Define the country before judging the traffic
A campaign target, an observed device location, a customer's declared residence and a commercially eligible market are different concepts. Give each its own field in the investigation. Do not rename them all “country” and expect the totals to reconcile.
Google explains that geographic location can be estimated from multiple signals, including network and device information. That makes the meaning and timing of a location observation important. An observation associated with advertising is not proof of citizenship, permanent residence, payment eligibility or the location of a later purchase.
The following register is a proposed working tool, not a Google reporting specification. Populate it from the actual systems available to your team.
Country definition Question it answers Evidence owner Common comparison error --- --- --- --- Campaign target Which markets did we configure for this campaign? Media operator Treating a campaign name as the saved configuration Reported geography Where does this system place the measured activity? Analyst Comparing reports with different dimensions or dates Customer market Which market does the product associate with this customer? Product or operations Assuming a profile field describes the ad interaction Commercial eligibility Can this customer receive and pay for the intended service? Business owner Treating an install as a serviceable paying customer
Write down when each definition is assigned and whether it can change. A customer who travels, updates a profile or makes a later purchase may appear in different slices without any one report being internally broken. That is a possible explanation to test, not an excuse to dismiss a material discrepancy.
Do not import a Search campaign fix without checking its scope
The familiar instruction to switch from “Presence or Interest” to “Presence” appears throughout generic location-targeting advice. Its relevance must be established for the campaign being investigated.
Google's advanced location options documentation names Search, Display and Demand Gen in its applicability statement. That statement does not establish that every App campaign subtype exposes the same controls. Inspect the current App campaign settings and applicable help before promising a specific switch. A missing control is a reason to verify scope, not to recreate the campaign through an unrelated setup route.
For App campaigns, Google documents that advertisers supply locations and languages while the system assembles and serves ads across multiple properties. Its App campaigns overview also advises against major geographic changes during learning. Neither point means a known incorrect target should be left indefinitely. It means the correction needs a recorded reason and an evaluation plan.
Keep the campaign type and subtype beside every screenshot. A settings example from Search, a Smart campaign or another advertiser's account is not evidence of the controls present in your app account today. When documentation and the interface do not resolve the question, retain both and ask Google support about the precise campaign rather than inventing a workaround.
Separate a reporting mismatch from a market allocation problem
“Wrong country” can describe several different incidents. Use the smallest decision that fits the evidence.
Observed problem First check Decision that may follow --- --- --- An unintended location is saved in campaign settings Current configuration and change history Correct the confirmed setup error Analytics and acquisition reports show different country shares Definitions, attribution scope and observation dates Repair or document the comparison An intended country receives most of the spend Mature customer outcomes and market constraints Retain the mix or evaluate market separation Installs arrive but users cannot use the core service Product availability and eligibility checks Contain exposure and fix the commercial mismatch Material unexplained delivery remains after reconciliation Reproducible campaign-level evidence Escalate a bounded support investigation
These are investigation paths, not automatic findings. A campaign can contain both a setup error and a reporting difference. Resolve the demonstrated error without claiming it explains every remaining row.
If the campaign has stopped serving altogether, use the Google App campaign delivery diagnostic. A geographic investigation is narrower: it asks where the delivered activity belongs and what that means for the next decision.
Build one comparable evidence window
Export the current target configuration, exclusions where available, language settings, campaign identity, app identity and relevant change history. Preserve the evidence before editing. A screenshot of today's configuration cannot establish what was saved when last week's activity occurred.
Then specify the reporting window. Record the account time zone, event dates, extraction time and whether the report describes ad interactions, attributed installs, first opens, active users or purchases. Keep the campaign identifier rather than relying on a naming convention that may have changed.
Use the same population wherever the tools permit. An all-user product report includes people outside the paid acquisition cohort. A purchase report for existing customers cannot be compared directly with new-install geography. Where the populations cannot be aligned, label the limitation and avoid presenting the percentage difference as wasted spend.
Do not force an individual-level join where the available measurement does not support it. An aggregate reconciliation with an explicit unmatched category is more useful than a falsely complete customer map. The app conversion tracking investigation addresses the separate question of whether events reach the intended destination.
Before making an economic claim, state which denominator supports it. Spend per reported install, spend per eligible customer and spend per retained payer answer different questions. The geography audit should identify that distinction before anyone declares a country cheap or expensive.
Uneven country spend is not automatically an error
A multi-country campaign does not represent an agreement to divide its budget evenly. Google's App campaign network distribution guide suggests considering separate campaigns for country tiers or economic profiles when more even country spending is needed. It also advises consolidating campaigns that target the same app ID, location and language to avoid overlapping traffic.
Those recommendations create a practical tradeoff. Separate markets when they require different commercial decisions; do not multiply campaigns merely to make a reporting chart look balanced. A distinct budget owner, service restriction, acceptable acquisition cost or language-and-offer combination can justify evaluating separate control. A different country label alone does not settle the architecture.
Before splitting, explain what the new boundary will control and what evidence it may fragment. Identify the success event and expected evaluation conditions for each resulting campaign. Do not assume a smaller market can produce equally useful optimization evidence on the same schedule.
Also consider whether the apparent media problem starts in the product. If onboarding, pricing or the service promise is unsuitable for a market, a separate campaign does not repair that mismatch. The mobile app localization framework covers that wider commercial decision.
A fictional example: cheaper installs, fewer eligible customers
Consider a deliberately simplified example, not a Sharply Labs result or a market benchmark. Two intended markets each receive $6,000 of spend during a completed comparison window. Market A records 1,000 attributed installs and 100 customers who pass the company's defined service-eligibility check. Market B records 2,000 attributed installs and 60 eligible customers.
The observed cost per install is $6 in A and $3 in B. Spend per eligible customer is $60 in A and $100 in B. These calculations use the same stated spend but different outcomes. Neither establishes profit, incremental acquisition or the cause of the difference.
Now suppose some customers change their declared market after registration. The analyst should preserve the original cohort definition and separately describe the later customer-market field. Reassigning customers silently would change the comparison while leaving the campaign expenditure fixed.
The next decision depends on what the evidence supports. If B's eligibility failure comes from an unavailable service, correct that mismatch before treating a lower install cost as success. If both markets are serviceable but their economics differ, evaluate whether campaign goals and market controls reflect the business decision. If outcome collection is incomplete, repair measurement before issuing a confident budget recommendation.
A geographic split might eventually be justified. The example does not prove that it will improve acquisition costs. It demonstrates why an install-only comparison cannot authorize the decision on its own.
Correct in an order that preserves attribution of the change
First contain a verified exposure the business cannot accept, using the controls available and the account owner's authority. A known service or policy restriction should not wait for an ideal experiment. Document the reason and the evidence that made containment necessary.
Second correct demonstrated configuration errors. Record the before and after states, effective time and responsible owner. If several unrelated changes are made together, the team loses the ability to identify which one explains the next observation.
Third repair the reporting comparison. Keep the previous extraction so the correction can be reviewed. Reconcile campaign identity, population and geography definitions before judging the new period against the old one.
Finally evaluate structural changes such as separating markets. Define the decision event, spending exposure and observation conditions in advance. The App campaign budget-change framework explains why a spending allowance and an optimization target should be evaluated separately.
There is no universal percentage of out-of-market activity that this guide can declare acceptable. The business model, consequence of an ineligible user, available measurement and applicable obligations matter. A numerical tolerance without those inputs would offer precision without a defensible decision.
What an agency or support packet should contain
Ask for a compact incident record that another operator can reproduce. Include the affected campaign IDs, app, subtype, intended locations, configuration evidence, relevant edit times, date range, dimension definitions and the smallest export demonstrating the mismatch. Redact unnecessary personal information and credentials.
Separate verified findings from hypotheses. “This report places these attributed events in an unexpected country” is a finding about a report. “Google ignored our targeting” is a causal claim requiring additional evidence. “The traffic is fraudulent” requires a different investigation; geographic disagreement alone does not establish it.
For an agency review, request deliverables that can be inspected: a map of the competing country definitions, a list of demonstrated setup defects, unresolved measurement limits, a recommendation to retain or separate markets, and an acceptance plan with named owners. Detailed implementation scope should be agreed after the evidence is understood.
For platform support, ask a narrow question about the documented discrepancy and available controls. Keep the case reference and response with the incident record. Avoid repeated campaign recreation or untracked changes while asking support to investigate the original state.
Questions to settle before authorizing more spend
Can location targeting guarantee the customer's country?
No. Google describes location targeting as a best-effort process with no guarantee of complete accuracy. It also does not establish a customer's commercial eligibility. If exact eligibility matters, the product and business must define their own appropriate checks; this article does not provide legal or technical implementation advice for them.
Should every country have its own campaign?
Only evaluate that structure when the markets require separate decisions and the evidence can support those decisions. Keep a written reason for each boundary. A requirement for separate control differs from a preference for equal spend, and neither proves a performance improvement.
Should an unexpected country be excluded immediately?
Correct a verified prohibited or unintended target using applicable controls and proper authority. When only a reporting difference is established, preserve the evidence and investigate before treating the row as a targeting defect. Do not assume every exclusion option from another campaign type applies.
Scope the review around the decision
For established app teams deciding whether to repair geographic reporting, correct targeting or separate markets, Sharply Labs' mobile app growth work and paid acquisition services provide a relevant starting point.
Contact Sharply Labs with a redacted configuration record, campaign change history and the definitions behind the conflicting country reports. The initial conversation can scope the mismatch, responsible owners and validation work; a detailed audit is separately agreed. No geographic precision, platform outcome or acquisition-cost improvement is promised.