<aside> 📌
A commercial gap closed without shipping a worse screen. Competitors had Google Reviews and we did not. I designed the guest-facing solution, and flagged that adding it inline would break a confirmation screen that already had no clear hierarchy.
Built once, live in 4 products. The component was modular from the start, so Pay at Table, Open Check and the mobile app gained it with no additional design work.
Configuration for operators running many locations. I took on channel-level setup after a colleague left, and designed the status and error handling so a failed Place ID cannot go live unnoticed.
</aside>
<aside> 🧭
</aside>
A missing feature that kept costing sales
Despite Online Ordering letting guests order ahead for collection and delivery for over a decade, it still didn't have Google Reviews. Even worse, our competitors had it, and it kept coming up in sales conversations. That is the moment when a missing feature does the most damage, so after enough of those it arrived on the roadmap as a priority and I was assigned to come up with the guest-facing solution.
The scope might have seemed small and simple. Just adding a prompt on the confirmation screen after checkout. And a bit of configurability for operators to switch it on.
As the project kicked off, we faced a real dilemma, and that eventually led me to notice the massive UX debt that essentially blocked us from delivering a quality feature.
Initial Confirmation Screen in Online Ordering
Guest-facing design, then channel-level configuration
Initially I had full ownership of designing the guest-facing experience. When the initiative started, I was working alongside a colleague who owned the store-level operator dashboard. Later, when we realised the channel-level configuration was equally crucial to the feature, I was assigned to tackle this myself.
Exploring ideas with AI
There were tons of ideas. I considered a drawer over the confirmation screen, an inline section inside it, a pop-up modal and many more. Interestingly enough, while benchmarking competitors, we came upon a very specific flow that would show the link to Google reviews only after the user gives a positive star rating inside the portal. As this was the solution we were losing prospects to, we asked ourselves, should we do it too?
<aside> ⚖️
This was clearly a dark UX pattern and furthermore, a compliance issue. Filtering for happy customers before you show them the review link is review gating and Google prohibits it. That ruled out this and a few more similar concepts.
</aside>
This was also the part of my workflow where AI helped significantly. I used Figma Make instead of drafting the ideas myself, ending up with over 70 variants. Granted, most of them were rubbish and a good chunk of them was just the same idea in a different arrangement. That's fine. Past a certain volume you stop reading individual screens and start seeing which families of solution actually behave differently from each other. I ended up with about 4 of them worth considering.
Initial AI-generated concepts