After the Redesign: Retesting Your Cookie Banner Before It Costs You
Nobody redesigns a website to break its cookie banner, yet it happens constantly. The new theme loads scripts in a different order, the marketing team adds fresh embeds, a chat widget arrives, and the consent setup that worked for two years is silently firing trackers before anyone chooses anything.
The fix is not a full compliance audit. It is a disciplined retest in the first week after launch, plus a habit of treating the banner as part of the site that ships. This guide gives the sequence and the specific failures to hunt for.
Start from a truly fresh visit
Testing from your usual browser tells you nothing, because it carries your old consent cookie. Use an incognito window or a clean browser profile, and a device you have not used for testing before if you can. Load the homepage and watch the order of events: the banner should appear before trackers fire. Open the browser's cookie storage and list what exists before you touch the banner. Anything beyond the consent platform's own strictly necessary items is a finding. Repeat on a phone, where templates and lazy loading often behave differently, and on your most embedded heavy page, usually a blog post with video and social widgets.
Reject, then browse like a determined sceptic
Choose reject or the strictest option available, then browse widely: product pages, checkout or contact flows, search, embedded content, and any page with a form. After each type, check what cookies have appeared. Embedded maps, videos, and review widgets are the usual offenders, loading their own trackers on scroll or interaction. Also test what happens when you return the next day in the same profile: the banner should stay away, the choice should hold, and no new categories should sneak in. Record findings with the page and the action that triggered them; vague notes like some cookie on mobile are how fixes stall between teams.
Accept, and verify the other direction
Now clear everything and accept all. Analytics should register the visit, consent mode states should update on the same page, and advertising tags should behave as designed. Then open the preference centre and change one category: turn marketing off while leaving analytics on, save, and confirm the tags respect the split rather than treating it as a full rejection or a full acceptance. Partial consent is where platform and tag wiring most often disagree. Finally, find the cookie list page on your site and compare it against reality. A public list naming cookies you retired in the redesign, or missing the new chat tool, tells visitors and regulators that nobody is minding the store.
Make the retest part of every launch, not a rescue
The durable fix is procedural: add the banner retest to the launch checklist next to redirects and forms, with a named owner. Keep a one page note of how consent is wired on your site: platform, tag order, where categories are edited, and who to call. Redesign agencies rarely include consent in their scope unless asked, so ask, in writing, before the build starts. The cost of the retest is a few hours. The cost of discovering, a year later, that every consent record since the redesign was collected while trackers pre fired is a conversation nobody wants to have with a regulator or a client.
Related reading
These earlier guides cover connected questions in more depth: Consent Mode Basics for Small Sites: What It Does and What It Does Not, Top Cookiebot Alternatives Worth Switching To, and Consent on Webinar Registration Pages: Tracking Before the Form Is Even Sent.
Frequently asked questions
How soon after a redesign should I retest the banner?
Within the first week, and ideally on staging before launch. Blocking problems affect every visitor from day one, so early findings are worth far more than a thorough audit a month later.
What is the most common redesign break?
Script order. The new theme loads analytics or marketing tags before the consent defaults or the banner script, so data is collected before a choice exists. Embeds added during the redesign are the second most common.
My banner looks fine. Is that enough?
No. Most failures are invisible: the banner renders perfectly while blocking fails underneath. You have to inspect cookies and tag behaviour in fresh and rejected states, not just look at the page.
Who should own the retest, us or the agency?
You should own the checklist and the sign off, even if the agency runs the steps. Consent responsibility stays with the site owner after the agency has moved on.