Consent mode and broken tracking: the quiet reason paid ads look worse

What consent mode does

Google's consent mode passes your visitors' consent choices to Google tags. It uses consent types including `ad_storage`, `analytics_storage`, `ad_user_data` and `ad_personalization`. Your cookie banner records what the visitor agreed to; consent mode tells your Google tags what they may do as a result.

When this is set up well, your tags respect each visitor's choice and the ad platform still receives what it is permitted to receive. When it is set up badly, two things can happen. Tags can fire when they should not, which is a privacy problem. Or tags can be held back even when the visitor said yes, which means conversions go missing from your reports.

How tracking usually breaks

These are the patterns worth checking first. Each one is a hypothesis to test, not a diagnosis:

• The banner and the tags disagree. The banner records consent, but the setting never reaches the tags, so they behave as if nobody consented.
• A theme or app update removed a tag. Common on hosted shop platforms after a redesign.
• Duplicate tags. The same conversion tag installed twice, by an app and by hand, inflating some numbers and confusing others.
• The thank-you page changed. The conversion tag was attached to an old order confirmation page that no longer exists.
• Consent defaults are missing. Tags have no instruction before the visitor makes a choice, so behaviour is unpredictable.

A simple check with Tag Assistant

Tag Assistant is Google's debugging tool. Used with Tag Manager's preview mode, it lets you inspect which tags fired and in which order. A basic run looks like this:

1. Open tagassistant.google.com and connect it to your site.
2. Visit the site as a new visitor and decline cookies. Note which tags fire.
3. Repeat as a new visitor and accept cookies. Note which tags fire now.
4. Go through a real purchase or enquiry journey to the confirmation page, and check that the conversion tag fires there.
5. Write down anything that surprised you, with screenshots.

If the tags look the same whether you accept or decline, or the conversion tag never fires on the confirmation page, you have found something worth fixing before you judge the ads.

Looking at rivals from the outside

You can see which tags a public website loads by looking at its page source or using a browser tag inspector. This is useful context. If every rival has a working consent banner and the major ad tags, and you are missing one, that is a gap. Keep it to public pages, respect each site's robots.txt if you automate anything, and treat what you see as an observation. You cannot see how well a rival's tracking works inside their ad account.

Fix the measurement before you move the money

If the tracking is broken, every channel decision built on it is shaky. Cutting a campaign because it shows few conversions, when the conversions are simply not being recorded, can remove something that was working. The order matters: check tags, fix what is broken, let the reports settle, then judge the ads.

Hypothetical example: a skincare shop sees reported conversions fall after a theme update. A Tag Assistant run shows the purchase tag is no longer on the new confirmation page. Fixing that one tag changes the story completely, without touching a single ad.

Make checking a habit, not an emergency

Tracking rarely breaks on its own. It breaks when something else changes: a new theme, a new app, a new cookie banner, a new checkout, a developer tidying up code. So the best time to check is straight after any of those changes, not when the reports start looking odd.

A simple routine helps:

• Keep a short written list of the tags that should fire on your key pages, and where.
• After any site change, repeat the accept-and-decline test above on the homepage, a product or service page and the confirmation page.
• Keep the screenshots with a date, so you can compare next time.
• When reported conversions move sharply, check the tags before changing any campaign.

If you work with an agency or freelancer, ask them to run the same check and share the screenshots. It is a fair question to ask anyone who manages your ad spend.

Next step

Our guide to tracking and measurement checks covers this in more depth, and the tracking tag regression watch recipe shows how to catch breakages after site changes. The own-ads audit loop is in build and not on sale. Start with the free Insight Report.