A well-designed mobile location consent flow explains the value before asking for access, records a precise consent state, and gives people a practical way to change their choice. This guide shows how to plan that flow across iOS and Android, estimate its operational impact with a repeatable model, and test denial, revocation, deletion, accessibility, and regional requirements before release.
Overview
Location permission UX is not a single dialog. It is a sequence of product decisions: why location is needed, when it is requested, what happens after a choice, which data is retained, and how a person can withdraw permission. For proximity marketing, location-based advertising, store visit measurement, or geofencing marketing, the permission experience should be tied to a clear user benefit rather than presented as a technical requirement.
A privacy-first flow has four properties:
- Purpose limitation: the app explains the specific feature or service that needs location. “Improve your experience” is less useful than “Show nearby pickup options while you use this feature.”
- Choice: the user can continue when location is optional, and can revisit the decision later.
- State accuracy: the app distinguishes not requested, denied, limited or approximate access, granted access, restricted access, and unavailable access instead of treating every non-grant as the same.
- Data minimization: the SDK and backend collect only the precision, duration, and event types needed for the stated purpose.
Platform behavior changes over time, and exact permission labels and settings paths can vary by operating-system version, device configuration, and implementation. Treat the platform comparison below as a planning model, then verify the current Apple and Android documentation and your SDK's release notes before shipping.
| Planning state | iOS implementation consideration | Android implementation consideration | Product response |
|---|---|---|---|
| Not requested | The app has not initiated the relevant system request. | The app has not requested the relevant permission set. | Explain the feature first; request only at the point of need. |
| Granted, with precision or scope limits | Access may include choices about precision and foreground or background use, depending on the request. | Access may include approximate versus precise location and foreground versus background behavior, depending on permissions and version. | Operate within the granted scope and disclose any reduced functionality. |
| Denied or declined | The app cannot assume a later request will behave like the first request. | The app should distinguish a denial from a temporary failure or unavailable provider. | Keep the core experience usable and offer an in-app explanation or settings route where appropriate. |
| Restricted or blocked by device or policy | Device controls, parental controls, enterprise policy, or other restrictions may affect access. | Device policy, profile controls, or system settings may affect access. | Show a non-blaming diagnostic message and provide an alternative workflow. |
| Revoked or changed in settings | The app must re-check authorization when it returns to the foreground. | The app must re-check authorization rather than relying only on a cached result. | Stop dependent collection, update the consent state, and explain the effect on features. |
Do not use a permission grant as a blanket marketing consent. Location access, marketing communication consent, analytics consent, and consent for personalized advertising can be separate decisions. Your consent-management design should represent those purposes independently.
How to estimate
Before implementing a flow, estimate its likely operational impact with a simple decision worksheet. The goal is not to predict a universal opt-in rate. It is to compare alternatives using the same assumptions and identify where a change in UX, collection scope, or fallback design affects the product.
Use these measures for each release or experiment:
- Eligible users: the number of active users who reach the feature that genuinely needs location.
- Prompt reach: eligible users who see the in-app explanation or pre-prompt.
- System-request rate: users who proceed from the explanation to the operating-system request.
- Grant rate: users who grant the required scope, divided by users who receive the system request.
- Usable-consent rate: users whose granted state is sufficient for the feature, divided by eligible users.
- Fallback completion rate: users who complete the non-location alternative after declining or being unable to grant access.
- Withdrawal response time: the time between a changed permission state and the app or backend stopping dependent processing.
A basic estimate is:
Usable consenting users = eligible users × prompt reach × system-request rate × grant rate × scope-sufficiency rate
For example, use an illustrative planning case of 10,000 eligible users, 80% reaching the explanation, 70% continuing to the system request, 45% granting access, and 90% receiving a scope sufficient for the feature. The estimate is 10,000 × 0.80 × 0.70 × 0.45 × 0.90 = 2,268 usable consenting users. These are assumptions, not a benchmark or forecast. Replace them with your own observed data after a controlled release.
Also estimate privacy operations separately:
- Consent records: eligible users × number of purposes recorded per user.
- Deletion workload: deletion requests × systems and vendors that must be checked.
- Support workload: users encountering denied, approximate, restricted, or revoked access × expected contact rate.
- Data exposure: event volume × retention period × number of systems receiving the event.
This framing prevents a narrow focus on consent conversion. A flow that produces more grants but collects unnecessary precision, obscures withdrawal, or creates an unmanageable deletion path is not a privacy-first improvement.
Inputs and assumptions
Document the following inputs in a decision record before development:
- Purpose: identify the exact feature, such as nearby store discovery, navigation, a proximity notification, or an opted-in location analytics function.
- Required scope: state whether the feature needs one-time, foreground, or background access, and whether approximate location is sufficient.
- Data lifecycle: specify what is collected, whether raw coordinates are needed, where processing occurs, who receives the data, and how long each data class is retained.
- Identity link: document whether location events are linked to an account, device, campaign identifier, or no persistent identifier. Avoid creating a profile when an aggregate event will do.
- Consent purpose: separate location permission from marketing, analytics, personalization, and advertising choices.
- Regional rules: identify the markets in which notice, consent, withdrawal, age handling, or vendor disclosures require additional review. Do not assume one global flow is appropriate everywhere.
- Fallback: define what the user can do without location, such as selecting a store manually, scanning a QR code, entering a postcode, or using a non-personalized experience.
Use a state model that can be understood by product, engineering, analytics, support, and privacy teams:
location_state: not_requested | explanation_seen | system_requested | granted_full | granted_reduced | denied | restricted | unavailable | revoked | deleted
purpose: feature_location | analytics | personalized_marketing
platform: ios | android
permission_scope: one_time | foreground | background | unknown
precision: precise | approximate | unknown
consent_version: version identifier
recorded_at: timestamp
updated_at: timestamp
source: app | settings_return | account_action
Store state transitions and purpose separately from raw location events. A revocation should trigger a clear transition and stop future processing that depends on the withdrawn purpose. Account deletion should remove or de-identify associated records according to the approved retention and legal process; deleting a local flag alone is not sufficient.
Copy templates
In-app pre-prompt: “Allow location for [specific feature]? We use your location while [context] to [clear benefit]. You can continue without it by [alternative]. We do not need location for [unrelated feature]. You can change this choice in Settings.”
When background access is genuinely required: “This feature works when the app is not open because [specific user-facing reason]. We will use location for [purpose], not for [unrelated purpose]. You can pause or withdraw access at any time.”
After denial: “Location is off, so [feature limitation]. You can [alternative action]. If you want to enable it later, use [settings path or in-app control].”
System prompt guidance: do not imitate or rewrite the operating-system dialog as if it were your own. Let the platform request provide the system-controlled choice, and ensure your preceding explanation matches the permission scope actually being requested.
Worked examples
Example 1: Nearby store discovery
A retail app wants to show nearby stores when a user taps “Find a store.” The feature can work with approximate location or manual postcode entry. The least intrusive design is to request access at the feature moment, explain the benefit, accept a reduced-precision state if it is sufficient, and offer manual search immediately. The decision record should say that background access is not required and that the marketing team cannot reuse the feature's location permission as consent for location based advertising.
Example 2: An opted-in proximity notification
An event app wants to notify a user about a saved venue while the app is not open. The team should first establish that the notification is an expected, user-requested feature. The flow should explain why background access is needed, record the relevant notification and location choices separately, and define a quiet or pause control. QA should test what happens when the user grants foreground access but not background access: the app should describe the limitation rather than repeatedly prompting.
Example 3: Consent withdrawal and account deletion
A user turns off location in device settings and later deletes the account. On the next app launch, the SDK should report the changed state, dependent collection should stop, and the backend should process the deletion request across the documented data stores and processors. Analytics dashboards should distinguish “permission revoked” from “no event received”; otherwise a measurement gap may be mistaken for a technical outage.
For event names and transition logging, align the implementation with a documented taxonomy such as Location Analytics Event Taxonomy. For SDK architecture, see Designing a Proximity Marketing SDK. These links should complement, not replace, current platform and legal review.
When to recalculate
Revisit the flow before every app release that changes location features, SDKs, identity links, retention, or vendor processing. Also review it before a new regional launch, a consent-rate experiment, a geofencing or proximity marketing campaign, an account-deletion redesign, or a privacy audit.
Run this QA workflow on supported device and operating-system combinations:
- Install fresh and confirm that no location request appears before the relevant feature is used.
- Read the pre-prompt with a screen reader, larger text, high contrast, and touch targets that do not depend on color alone.
- Test grant, denial, reduced precision, foreground-only access, and any supported background path.
- Close and reopen the app, return from system settings, and verify that authorization is re-read.
- Test restricted, unavailable, offline, and provider-error states without trapping the user in repeated prompts.
- Withdraw access, confirm that dependent collection stops, and check that the backend records the transition.
- Delete the account and verify the documented deletion or de-identification process across relevant systems.
- Repeat regional variants for notice, consent, marketing preferences, age handling, and vendor disclosures.
- Compare observed funnel inputs with the estimate: eligible users, prompt reach, request rate, grant rate, scope sufficiency, fallback completion, and withdrawal response time.
Refresh the implementation notes after major operating-system or SDK releases and conduct a broader compliance review at least twice a year. Keep the previous consent copy, purpose description, state model, data-flow diagram, test results, and approval decision together. That record makes future recalculation faster and helps product teams improve consent without treating people or their location data as conversion inventory.
For the wider privacy-first approach, see Privacy-First Location Marketing and Consent-Based Identity Resolution for Location Marketing.