Cookie consent testing by country and US state

Banner variant served

Which banner, or none, the market received. The first assertion: is this the configuration we intended for this country or state?

Cookies set before interaction

Name, domain, expiry and whether each is classed as essential. Non-essential cookies here are the headline finding in the EU.

Third-party requests before interaction

Domains contacted before the visitor chose. A tracker request without a cookie can still be a problem; record both.

State after reject

Cookies and requests after a reject click and after a reload. Shows whether the choice is applied and persists.

State after accept

Confirms tags work when consent is given, so the fix for one bug does not break measurement.

First-layer options

Whether reject is offered next to accept, and whether boxes are pre-ticked on the settings layer.

GPC handling

Whether a request with Sec-GPC: 1 changes what is set and shown, and whether the site displays that the signal was honoured.

Exit country, city and timestamp

Evidence of where the test ran from, so a finding can be reproduced and explained.

1. 01

   Define the matrix: pages × markets × states

   List the templates that matter (home, product, checkout, article), the markets you serve, and the three consent states. Add a GPC variant for US markets that require opt-out signals to be honoured.

2. 02

   Run each cell in a fresh context through the market's exit

   Open a new browser context with no stored state, held on one sticky exit for the whole cell so the page, banner and reload all come from one visitor.

   Pre-consent and post-reject capture, Playwright26 linescopy

   from playwright.sync_api import sync_playwrightSERVER, PASSWORD = "http://proxy.arcproxies.io:8000", "PASSWORD"def run_cell(url: str, country: str, city: str | None, cell: str, gpc: bool = False):    user = f"USERNAME_pool-arcfast_country-{country}"    if city:        user += f"_city-{city}"    user += f"_session-{cell}_sesstime-10"    with sync_playwright() as p:        browser = p.chromium.launch(proxy={"server": SERVER, "username": user, "password": PASSWORD})        ctx = browser.new_context(extra_http_headers={"Sec-GPC": "1"} if gpc else {})        page = ctx.new_page()        requests = []        page.on("request", lambda r: requests.append(r.url))        page.goto(url, wait_until="networkidle")        before = {"cookies": ctx.cookies(), "requests": list(requests)}        page.get_by_role("button", name="Reject all").click()  # adapt per banner        page.reload(wait_until="networkidle")        after = {"cookies": ctx.cookies(), "requests": list(requests)}        browser.close()        return before, after# California, with and without the GPC headerrun_cell("https://www.example.com/", "us", "losangeles", "ca-gpc", gpc=True)run_cell("https://www.example.com/", "de", None, "de-base")

3. 03

   Classify what you recorded

   Map every cookie and request domain to a purpose using your consent platform's categories. Anything non-essential before consent, or still present after reject and reload, becomes a finding with the market attached.

4. 04

   Run it on a schedule and on release

   Run the full matrix weekly and a reduced matrix on every release that touches templates, tags or the consent platform. Diff against the last run so a new tracker shows up as a change, not as noise.

Prices are list prices in EUR, excl. VAT. Volume discounts apply automatically at checkout — full ladder on the pricing page.