Analytics without cookies: what you can measure and what you cannot
How cookieless analytics works, which questions it answers as well as the cookie-based kind, and which it cannot.
The short answer
Cookieless analytics measures a website by counting rather than identifying. It records each page view with the page, the referrer, the country, the device type and the time, groups views into visits within a day using a short-lived hash that is thrown away, and stores nothing on the visitor’s device. Because no identifier is set or read, it needs no consent banner, it counts every visitor rather than the minority who accept, and its script is a fraction of the size of the cookie-based tools. It answers the questions a business site actually asks: how many visitors, from where, which pages they read, which actions they took, and what changed after a change was made. What it cannot do is follow one person across days or devices, build audiences for advertising, or attribute a purchase to an advertisement clicked weeks earlier. For an ordinary business website those losses are theoretical and the gains are real, which is why it is the right default, with the ad platform’s own consent-gated measurement added only where significant advertising justifies it.
What you can and cannot measure
| Question | Cookieless analytics | Cookie-based analytics |
|---|---|---|
| How many visitors and page views, by day | Yes, for every visitor | Yes, for consenting, unblocked visitors |
| Where visitors come from: search, social, referrals, campaigns | Yes | Yes |
| Which pages are read, how long, where visitors leave | Yes | Yes |
| Which countries and devices | Yes | Yes |
| Goals: form sent, purchase, download, click | Yes, with the visit’s source | Yes |
| Effect of a change: before and after | Yes | Yes |
| Returning versus new visitors over weeks | No | Yes, for consenting visitors |
| One person’s path across several days or devices | No | Partly |
| Audiences for retargeting | No | Yes |
| Attribution to an ad clicked days ago | No | Partly, with declining reliability |
| Consent banner required | No | Yes |
| Share of visitors measured | All | The consenting, unblocked share |
Setting it up well
- Choose a cookieless service hosted in a suitable region, with terms that fit European requirements.
- Add the light script or the server-side integration; verify in the browser that no cookie is set.
- Define goals: form submissions, key page views, downloads, purchases, outbound clicks that matter.
- Tag campaigns consistently so sources are meaningful.
- Remove the cookie-based tool and its consent category, and update the privacy statement.
- Build one dashboard with the numbers the business reviews monthly.
- Add real-user performance monitoring alongside, also without identifiers.
When cookie-based measurement still earns its place
A business running substantial paid advertising benefits from the ad platform’s own measurement, gated behind consent, for optimisation and attribution the platform needs. A product with logged-in users measures behaviour through its own application, where the user is known and the terms cover it. Both sit alongside cookieless analytics for the site as a whole. For everyone else, the cookie-based tool is a banner, a heavy script and partial data in exchange for reports the business does not use.
What this means for you
Cookieless analytics answers the questions a business website needs answered, for every visitor, with no banner and a light script, and gives up only cross-day tracking and advertising audiences. Make it the default, define goals, tag campaigns and review one dashboard monthly. Add the ad platform’s consent-gated measurement only if substantial advertising needs it, and let the rest of the site measure honestly and completely.
Frequently asked questions
How can analytics work without cookies?
By counting rather than identifying. Each page view is recorded with the page, the referrer, the country from the network address, the device type and the time, and visits are grouped within a day using a short-lived, non-reversible hash that is discarded. No identifier is stored on the device, so nothing is read back later and no consent is needed. The result is accurate counts and trends without a profile of anyone.
Is the data less accurate?
More accurate in one important way: cookie-based analytics only sees visitors who accept the banner, which on many sites is a minority, and is blocked by browsers and extensions for many more. Cookieless analytics counts every visitor. Less accurate in another: a person returning next week counts as a new visitor. For understanding how a business site performs, complete counts matter more than returning-visitor precision.
What about conversions from advertising?
Goals on the site, such as form submissions and purchases, are counted with the source of the visit that produced them, so campaign performance within a visit is visible. What is lost is attribution across days and devices, and audience building for retargeting, both of which depend on identifiers. Businesses that run significant advertising can add the ad platform's own consent-gated measurement for that purpose while keeping cookieless analytics for everything else.