A cookie banner appearing on an OpenCart page does not prove that consent signals reach your tags. Test what happens before the visitor chooses, after each choice and when that choice changes. The objective is to verify the implementation, not simply the banner’s appearance.
Google Consent V2 provides a configurable consent interface for OpenCart, including multilingual and multistore settings. If you use the consent functionality in Google Marketing Tools instead, identify which component owns consent and avoid running two independent controllers for the same page.
Identify the consent controller and tag owners
List the module responsible for the interface, the place where default consent is set and the code responsible for updates. Then list GTM containers and any analytics or advertising tags installed directly in the theme. A correct GTM setup cannot control an unrelated script that ignores the agreed consent process.
Use a test environment or a separate test stream when possible. Open a clean browser session so a previously saved choice does not hide the initial behavior. Include the product, cart and checkout routes, not just the homepage. Some themes use different layouts on those pages.
Check the four Google Consent Mode v2 signals
Review analytics_storage, ad_storage, ad_user_data and ad_personalization. They represent different permissions; one “marketing” label in a banner may update more than one signal. Confirm how your interface categories map to them rather than assuming every toggle is independent.
Follow Google’s implementation documentation when checking defaults, updates and tag behavior. Default consent should be established before relevant tags need it. Consent Mode is a technical signaling mechanism; your privacy requirements and the text shown to users need their own review.
Run a repeatable choice matrix
| Test | What to inspect |
|---|---|
| New session, no choice yet | Default state and the order of tag execution |
| Refuse optional categories | Updates reflect refusal without an unexplained grant |
| Accept analytics only | Analytics and advertising signals remain distinct |
| Accept all configured categories | Updates match the interface choice |
| Change an earlier choice | New signals replace the earlier consent state |
| Visit another page | The saved choice is handled consistently |
Use Tag Assistant’s consent information and the network panel together. Do not decide that any request to Google proves consent was ignored: behavior depends on whether your implementation uses basic or advanced mode. Compare the request and storage behavior with the mode you intentionally configured.
Test storage, navigation and multiple stores
After choosing, navigate to a product, cart and checkout page. Check that the correct choice persists for the intended scope and that a new page does not briefly reset it to an inappropriate state. Test a return visit as well as a fresh session.
For multistore installations, check each hostname and language separately. Cookie scope and translated interface settings can differ. An English banner working on one store does not establish that a second store sends the same updates. Record which store and language were used for each result.
Watch for duplicate banners and stale assets
Conflicting controllers can issue successive updates that make debugging confusing. Common places to check are a dedicated consent extension, a tracking extension’s built-in banner, a theme script and a GTM template. Keep one documented owner for the consent interface and a deliberate integration with the tags.
After changing settings, clear only the relevant application and browser caches, then repeat the clean-session test. Check the version of the script actually delivered to the browser. A configuration screen can show a new setting while cached JavaScript still applies the previous one.
Keep evidence that someone else can repeat
Save a short test record with the date, store, language, browser, implementation mode and observed states. Include a sequence of actions rather than a single screenshot of the banner. Re-run it after tracking, checkout or theme changes. If purchases also look duplicated, investigate that separately with our GA4 purchase-event checklist; consent testing should not be used to compensate for an overlapping event sender.