Cookie banner: how many visitors do you actually lose?

What the industry actually reports

CMP providers (Consent Management Platforms) like Cookiebot, Usercentrics and OneTrust regularly publish aggregated consent rates for their customers. The figures vary widely between industries and countries, but the general patterns are consistent:

• News sites and media: 60 to 75 percent accept. High figures because the dialog is familiar and the content is free-for-data.
• E-commerce: 50 to 65 percent accept. Depends heavily on how clearly the reject button is exposed.
• SaaS and B2B: 35 to 55 percent accept. Lower because visitors are often more privacy-aware and used to developer tools.
• Public sector and newsletters: 25 to 40 percent accept. People seeking government services usually click "reject" right away.

Which means that if you run a typical Swedish SMB site you miss 30 to 60 percent of your visitors in your web analytics. You pay to buy traffic via Google Ads, LinkedIn or SEO, and a large part of the traffic that actually lands on your site becomes invisible.

Worked example: what it costs concretely

Assume you have a B2B site with 10,000 unique visitors per month and a typical consent rate of 45 percent. In your GA4 report you then see roughly 4,500 visitors. The other 5,500 are there, you just do not see them.

The consequences are not abstract:

• False conversion rates. If you have 100 purchases/month you calculate wrong: GA4 says 100 / 4,500 = 2.2 percent. Reality is 100 / 10,000 = 1 percent. You are chasing the wrong KPIs.
• Incorrect channel attribution. Visitors who reject cookies more often come from types of traffic (organic, direct) you now underrepresent, while others (paid social, paid search) show up disproportionately. Your ad budget is steered by false data.
• Confusion in management reports. When the marketing department reports 4,500 visitors and the sales team talks about 80 leads, the conversion formula becomes completely wrong.
• Worse SEO feedback. You cannot see which of your ranking pages actually convert for 55 percent of the traffic. You optimize blindly.

Why not just "improve the banner"?

People have tried. The most common tactics and their problems:

Pre-checked boxes

Illegal since 2019 (the Planet49 ruling). You cannot have "accept" pre-ticked. The Court of Justice of the EU and IMY have been consistent.

Dark patterns (making reject hard to find)

You temporarily take the risk that a customer reports you to IMY for undue pressure. IMY’s guidance and the EDPB guidelines from 2022 explicitly say that the reject button should be as easy to find as the accept button. Fine levels are up to 4 percent of annual revenue.

Cookie wall (no access without consent)

Illegal in almost all EU countries. It is no longer "free" consent if the alternative is to lose the service. German and Danish supervisory authorities have ruled this out.

Reduce the number of trackers

Actually works, but marginally. If you reduce from 12 third parties to 6 your consent rate rises by about 5 to 10 percent. Not enough to solve the underlying problem.

The solution that really works: no banner at all

If none of your trackers require consent, you need no cookie banner at all. GDPR and the ePrivacy directive require consent for cookies and similar techniques that store information on the visitor’s device or create identifiers that track them across sites. Web analytics that avoids all of these categories requires no banner.

Concretely that means you need:

• No cookies, no localStorage, no sessionStorage on the tracking side.
• No IP address stored, no fingerprinting, no cross-site identifier.
• No US transfer that triggers additional information requirements.

This is the entire reason for existence of cookie-free European web analytics services. Spårlös, Plausible, Fathom, Simple Analytics and similar services follow the above rules from the ground up. You can remove the banner entirely if web analytics was the only reason you had it.

When you still need the banner

Unfortunately switching web analytics is not always enough. You still need a banner if you use:

• Facebook/Meta Pixel or LinkedIn Insight Tag for remarketing.
• Google Ads conversion tracking.
• Heatmap tools like Hotjar, Microsoft Clarity (in session mode), Mouseflow.
• A/B testing tools that store visitor segmentation in cookies.
• Live chat with conversation history linked to cookies.

The strategy for many then becomes two-part: use cookie-free web analytics for the basic insights (page views, sources, journeys), and collect consent only for specific tools where it really adds value (remarketing to purchase-ready segments, for example).

Tools to combine to skip or minimize the banner

• Web analytics: Spårlös, Plausible, Fathom, Simple Analytics, Matomo in cookie-free mode.
• Heatmaps without cookies: Microsoft Clarity in cookieless mode (it exists), or your own server-side measurements.
• A/B testing without cookies: server-side experiments via tools like GrowthBook (open source, self-hosted) or Statsig (cookieless mode).
• Conversion tracking without cookies: server-side tracking to Google Ads via Enhanced Conversions (does not require browser cookies).

The worked example again, after the switch

Back to the site with 10,000 unique visitors per month. You switch to Spårlös and remove the banner. Now all 10,000 visitors show up in your reports. Concretely:

• Conversion rate drops from the (incorrect) 2.2 percent to the (true) 1 percent.
• Channel distribution becomes more realistic: organic goes up, paid social goes down.
• Marketing and sales can start talking about the same traffic figures.
• You may discover that 30 percent of your visitors come from LinkedIn rather than Twitter, because LinkedIn visitors more often rejected cookies before.

Within 30 days you have a completely honest picture of your traffic for the first time since the ePrivacy directive was implemented in 2018.

How to start

If you want to test before removing anything:

• Create a free account on Spårlös.
• Install the tracker alongside GA4 for 48 hours.
• Compare the total visitor count. The difference is roughly what you lose to the cookie banner today.
• If you decide to: remove the GA4 code, remove the banner, update the privacy policy. Done within an afternoon.

Want to read the deep dive: why Swedish companies switch, or the comparison page for a side-by-side walkthrough. Pricing is at /en/pricing, from 59 kr/month including VAT.