Building a Secure Omnichannel Experience for a Top 100 Global Beauty Brand

Solution

The client offers an in-store consultation experience where customers can use digital beauty screens to explore product options, such as hair color, foundation shade, color type, and related product selection scenarios. The business needed this assisted discovery to continue after the customer left the store screen.

Instead of letting the recommendation context disappear at the end of the session, the next step was to bring the same product set into an ecommerce platform, where customers could review the recommendations and decide what to do with them later.

Relying on our ecommerce web development services, Expert Soft implemented the flow inside the client’s SAP Commerce Cloud and Spartacus context. The kiosk generated a QR code that opened an ecommerce page with the recommended products. From that page, customers could view the list of chosen items, share it by email without saving it as regular persisted data, or add items to a dedicated wishlist type separate from the default wishlist.

Security Concerns

The first technical idea from the kiosk side was to pass product IDs and tool categories directly through URL query parameters. Expert Soft challenged that approach because it weakened the integrity of the recommendation flow. A customer could modify the URL and make arbitrary product combinations appear as if they came from the in-store consultation.

The issue was not sensitive-data protection. The token data itself was not sensitive, but the business needed the recommendation set to remain trustworthy. If the ecommerce page accepted edited URL parameters as official kiosk output, the customer-facing journey would be easy to manipulate and harder to control.

To solve this proportionately, Expert Soft proposed and implemented a JWT-based flow. The kiosk generates a JSON Web Token containing the product IDs and required tool-related metadata. The QR code carries the ecommerce URL together with the token, and the back-end validates the token with a shared secret before accepting the recommendation data.

When the token is valid, the ecommerce page renders the kiosk-generated product recommendations and allows the supported follow-up actions, including email sharing. When the token is invalid, the back-end rejects the data and blocks the flow. The client’s security team reviewed the design and approved it as sufficient for this use case.

This gave the brand a simple customer journey without treating the QR link as an uncontrolled source of truth. Back-end validation preserved the integrity of the in-store recommendation set while still letting customers move naturally from assisted product discovery to ecommerce actions.