How it works

Age verification that is
simple, fast and compliant

Klikage combines three verification levels to fit every situation, without ever compromising your customers' privacy.

① In-browser age estimation

On reaching a protected page, a window invites the visitor to turn on their camera. Age estimation is computed locally, in their browser, using face-api.js (TensorFlow.js). No image is sent to our servers: only the verdict (adult / minor) is used. It's free and unlimited.

② Binding to the customer account

Once verification succeeds, the validated age is saved on the customer account. The visitor never uses the camera again — even from another device. Customers already registered on the shop at install time are considered adults by default.

③ Enhanced ID verification (KYC)

For cases where facial estimation is not enough (doubt, strong legal requirement), Klikage offers ID verification carried out by our approved KYC partner, Didit. We don't hide who handles these checks: it is Didit, a specialist in digital identity, that collects and checks the document. GDPR compliance of this step is Didit's responsibility as a processor: no document passes through your shop or our servers. This option consumes prepaid credits.

Manual ID document validation

You can also handle ID verifications yourself, without any external provider and without using credits. This is useful when a customer has no camera, or if you prefer to keep these checks in-house. Here's how it works:

  1. The customer clicks "Request manual validation" and fills in a form, sending the documents you have chosen (name, email, phone, ID), after ticking a GDPR consent box.
  2. The request lands in the module back-office (Customers › Klikage — Validations menu) and you are notified by email.
  3. You review the document, then validate or reject in one click. Validation records adulthood on the customer's account.
  4. The ID document is automatically destroyed once the request is handled — you keep no sensitive document.
The age threshold and the margin of error. Facial estimation has a margin of around 4 to 5 years. For a sale banned to under-18s, setting the threshold to 23 offers a good safety / smoothness trade-off: adults close to 18 are not wrongly blocked, while keeping a comfortable margin. You set this threshold in the module.