Untangling Woolworths and Rewards logins

The account model did not match customer expectations

Customers could create accounts for Woolworths and Rewards using the same credentials without the accounts being linked. Many people understood the brands as part of one service and expected that relationship to carry through to login.

The interface added more friction. Rewards sign-up could move customers out of the Woolworths app into a browser. The project account describes a lengthy journey, a form spread across more than ten pages and email passcodes that could arrive after they had expired. Customer care was handling substantial demand for help with account access.

Agree on what needed to change

The work began with existing login problems and expanded to a single sign-on that could eventually support other Woolworths businesses and partners. I worked with senior stakeholders, product managers and technical leads to align the objectives: resolve immediate customer issues, reduce the need for support and establish a path towards shared sign-on.

We separated the work into two releases so the immediate improvements could be delivered without waiting for the whole identity experience to change.

Map the account states before designing screens

The scenario map covered the different combinations of account and login state the experience needed to support. A decision tree then made the key customer actions easier to discuss.

These maps gave design, product and engineering a common starting point. They helped us identify where an apparently simple login screen depended on the customer’s existing relationship with either service.

Develop the design with the team

Sketching workshops gave team members a way to put their assumptions into a proposed solution. Participants explained their ideas and voted on the options they thought would work best.

I brought those ideas into low-fidelity wireframes and refined them through stakeholder and technical discussions. Working with the UI designer, we then reviewed interaction detail, visual consistency and the relationship between existing and new patterns.

Test login within a grocery task

The final study described in the project account was an unmoderated, unguided usability test with five participants. The task was to buy groceries, rather than to complete a login exercise in isolation.

That framing let us examine whether account access interrupted the activity customers had actually come to do. The account records the study’s design and participant count, but does not provide task-level results or a measured conversion change.

What shipped and what was deferred

The broader One Login project was deprioritised after testing. Release 1 had already delivered improvements to the immediate account-access experience:

• Code-based login was introduced for the app and website.
• Rewards sign-up could remain within the app instead of sending customers to a browser.
• Customers completing the relevant account-creation journeys could stay in the app and be signed in automatically.

The retrospective reports fewer routine login-related support calls and less negative app-store feedback. It does not include the underlying counts, comparison period or attribution analysis, so these remain qualitative observations rather than quantified results.

Reflection

The original retrospective noted that the broader programme had been reprioritised at the time it was written. This case study records that historical position; it does not assert the programme’s current status.

A gap in the follow-up was the lack of a check-in with customers who had previously left negative reviews. A stronger measurement plan would connect support-call categories, successful account access and follow-up customer feedback across a defined period. The delivered improvements and the proposed shared sign-on should also be evaluated separately.