Growing Table without spending restaurant trust

0.39%→3.73%

Table's share of reservations

156→1,246

Average daily bookings

1,182→19,181

First-time bookings, in the month after launch

15–30%

Increase in engagement, after Invite Friends

The reservation widget already gave guests a familiar way to book directly from restaurant websites. At the same time, Zenchef had invested in Table, a consumer app that offered restaurant discovery and a more persistent, personalized booking experience, but adoption remained low.

That created a strategic tension: the widget had the traffic Table needed, yet it was also a critical partner-facing conversion surface. Any promotion that interrupted booking could hurt completion, prompt restaurant pushback, or make the product feel like it was prioritizing Zenchef's app over a restaurant's guest experience.

Across Widget to Table and Invite Friends, I owned the B2C design work, working with product management, engineering, and restaurant partners. I mapped the existing journey, identified intervention points, designed and tested flows, and translated feedback into product iterations.

The work had several constraints:

• The reservation flow was high intent, so any extra step had to earn its place.
• Restaurant partners needed to retain confidence and control over an experience embedded on their own sites.
• Mobile and desktop required different solutions.
• Guest sharing needed to feel useful immediately while remaining privacy-conscious and technically flexible enough to support future CRM or loyalty capabilities.
• Mergers and acquisitions changed priorities and limited what could be carried through to later stages.

Promote Table before commitment, not after booking. The Table choice went in the widget's first view, where guests could choose to continue through Table or remain in the existing widget. This was the most natural decision point: it made the app's value visible before the booking flow became transactional, while keeping the original route available.

Treat mobile and desktop as different journeys. Initial designs included a desktop QR-code path to download the app. Feedback and adoption showed that it added too much friction for too little value, so the team retired the desktop version and focused the promotion on mobile, where the transition was more immediate.

Make booking coordination part of the product. Research showed that guests already shared reservation details externally. Invite Friends became an in-app continuation of the booking flow, letting a booker invite others and keep them informed without forcing the coordination work into messaging apps.

Design a data model that didn't force an unproven business model. The intended backend could retain invitee information for future bookings and enable future loyalty use cases. During the merger, the plan to expose that guest data to restaurants was dropped. The product retained the user value, simple coordination, without making privacy-sensitive data-sharing a prerequisite for launch.

Widget to Table surpassed its original adoption target. Table's share of reservations rose from 0.39% to 3.73%, average daily bookings grew from 156 to 1,246, daily sign-ups increased from 563 to 1,799, and daily active users rose from 2,485 to 6,028. First-time bookings increased from 1,182 to 19,181 in the month after launch. Invite Friends subsequently produced a 15–30% increase in engagement.

The work also made a partner-management gap visible: 5.63% of restaurants contacted support to opt out of the widget promotion. In retrospect, I would involve and prepare restaurant partners earlier, explain the customer and commercial rationale more clearly, and provide a more deliberate rollout and opt-out model. The product result was strong, but the rollout showed that a partner-facing surface needs change management, not only a well-designed user flow.