Rachael Hunt
← All work

Safe admin configuration

Making complex admin tools safe. Admins were editing booking policies but never switching them on. I made it obvious what's live, and safe to edit before publishing.

Role
Lead designer, with a product manager and the engineering team
Product
Clinic booking configuration, Best Practice Software
Market
Australian clinics using online booking
Methods
Hotjar, interaction design, content design

The problem

Clinics can show patients a booking policy (cancellation notice, no-show charges) during online booking. Hotjar showed configuration users editing their policies but never enabling them, so patients often never saw them.

A live policy

Booking policy settings marked Live, showing the clinic's policy and an Edit booking policy link

What I did

Clear state

A Live or Not in use label shows at a glance whether patients can see a policy, and admins can read it right there. New clinics get a sample policy, so they start well.

Booking policy settings marked Not in use, with a warning that no policy is displayed and the sample policy below
A new clinic's settings: clearly not in use, with a sample policy ready to edit.

Edit safely, then publish

Admins can save a custom policy partway through editing without it going live, and publish it explicitly when it's ready.

The Edit booking policy dialog: a rich-text editor with Save as a draft, Publish and Cancel
Drafts stay private until the policy is published.

Clear, explicit language

Every action says what it will do to the booking site, so admins understand the impact before they commit.

The Publish button's menu with Disable booking policy A confirmation: 'Disabling the policy means it's no longer visible when making a booking online', with Disable and Cancel
Disabling a policy asks first, and says what it means for patients.
A phone showing the clinic's booking policy in a dialog with an I understand button, during booking
What patients see: the policy, in plain language, before they book.

Outcome

This was part of wider work on the clinic booking configuration platform, which saved admins ~15 seconds on every doctor and appointment setup, used daily by 80+ clinics.

Try the prototype

How I'd use AI now

This project was done before AI tools were part of my process. If I did it today:

  • Offer clinics an AI-drafted policy written from their own settings (notice period, fees), in plain language, which they edit and publish themselves.
  • Check policies as they're written for unclear or inconsistent terms ("24 hours" in one place, "one day" in another) before they go live.
  • Summarise support tickets and Hotjar sessions about configuration to find the next most confusing setting.