Aesha Shah

Introduction to ResyOS Classic - the old platform

ResyOS Classic is the reservation inventory management system that Resy’s platform operated on from 2014 to early 2018. Within the iPad app, ResyOS Classic has three main views:

Discovery & Brainstorming

We decided to completely reimagine our inventory management system from both the back-end architecture as well as the front-end interaction model. In addition to gathering technical requirements from our CTO I also spent time understanding the benefits and gaps of our current system. To effectively transition approximately 2000 restaurants operating on the ResyOS Classic slot system, we would need to design a flexible, dynamic application that accounted for their many possible scenarios.

What restaurants like about Slots and ResyOS Classic

• Allow restaurants to have maximum control. They can configure their night exactly as they’d like to see it booked.

• Allow for granular control. Each slot on the slot grid can have a completely different configuration (min-max party size, cancelation policy or fee, paid reservation, the list goes on…).

• In one graphical view, restaurants can see exactly what’s booked and what is available for the shift.

Gaps

• Reservations and slots are not tied to a physical floor plan, so in turn they need to be manually plotted onto tables. Without an experienced maitre’d this can cause overbooking or underutilization of capacity.

• Lack of flexibility in changing the party size or time of a reservation during a shift.

• Inability to create exceptions to slot grids for special events like New Year’s Eve dinner, or Valentine’s day.

• Not a self-serve system, but rather requires restaurants to contact their Resy account managers to make changes to the setup.

Overall, a slot system caters well to very experienced maitre’ds and restaurants with complex configurations. But that is just a tiny sliver of the restaurants in the market.

I worked with Resy account managers to understand what information is required to construct a slots grid, and worked shifts at restaurants to witness how those grids were used in practice. Below is a grid from a sample restaurant and its adaptation to the ResyOS Classic slot grid.

I also conducted competitive analysis of OpenTable’s Electronic Reservation Book, OpenTable’s GuestCenter, Reserve, and Yelp Reservations.

I gathered 2 years-worth of feedback from account managers, worked shifts at restaurants of all sizes and demand levels both in-service and pre-service, as well as conducted contextual and retrospective interviews with maitre’ds, general managers, and restaurant owners. I also worked with our sales team to understand requirements for selling to new market cities.

ResyOS Fly, a Table-Based Inventory

While there were other table-based inventory systems out there, none of them took advantage of reservation optimization and automatization. Our goal with Fly was to not only provide inventory that was tied the physical floor plan, but also use Fly as a layer of intelligence to help guide restaurants in booking reservations that would optimize every shift with the following factors:

Inventory Mode - Fly, Slots, or Hybrid

We also recognized the value of the slot system’s granular controls, and therefore wanted to maintain this concept in ResyOS Fly. Restaurants can choose to operate entirely in slots, entirely in Fly, or choose to create specific slots and let Fly’s algorithm run around it. This allows restaurants to be very prescriptive for many possible scenarios, including high-demand nights, large parties, or during the prime-time of a specific shift.

The Timeline Inventory View Pre-Service and In-Service

Pre-service: view all reservations plotted on the timeline along with their reservation status - Reserved, Confirmed, or Left Message.
In-service: view all reservations plotted on the timeline with their status accurate to the floor plan - Partially Arrived, Arrived, Partially Seated, Seated, Appetizer, Entree, Dessert, Check Dropped, Finished.

Resy Fly Recommendations on the Timeline

While Resy Fly recommendations reveal what is bookable online on www.resy.com, the Resy Consumer apps, and the restaurant’s website - in ResyOS, Resy Fly acts an additional layer of guidance. A restaurant can trigger Resy Fly recommendations from the Timeline view by selecting a party size and entering booking mode.

The Spread of Covers

The spread of covers was not part of our original design for Resy Fly. Our prototypes of Resy’s new timeline view was received with cautious excitement from existing ResyOS Classic restaurants. On digging deeper with focus groups and shadowing maitre’ds, we learned that while we thought the slots grid was primarily used in the pre-service booking flow, in reality, restaurants often went back to it right before the start of shift as well as during in-service to see the “shape of the shift”. In pre-service team meetings, this “shape” was used to communicate to the front-of-house as well as the chef on when to expect spikes in activity and coordinate accordingly.

We observed that the maitre’d would print the spread report from ResyOS and have a couple of scribbles on it to highlight when VIPs were coming in, and well as reflect any changes in reservations since the time of printing. During service, the maitre’d would primarily operate on the ResyOS floor plan view, but at the end of every busy seating wave would switch to the slots view to prepare for the rest of the night.

We therefore added a live and dynamic version of the spread of covers report right into the iPad and make is accessible from both the timeline view and the floor plan view. The Spread of Covers shows the cover count of each reservation by time, party size, status, and VIP indicator. It also provides roll up information of active statuses as well as quick filtering capabilities.

Blocking Inventory

ResyOS Classic has the ability to block tables from the floor plan. This is used for partial buy-outs of an event or occasionally to hold tables for VIPs. While this block does not allow reservations to be plotted onto the blocked tables, it does not prevent them from being booked.
With ResyOS Fly, we created the ability to create full shift or partial blocks on a table from both the floor plan and the timeline view. Creating these blocks would occupy the table for the specified time range and prevent them from being booked.

ResyOS Fly Rollout

ResyOS Fly was a reinvention of itself and displayed Resy’s all-in approach on optimized inventory management. While our research and insights strongly suggested we had come up with a significant update for restaurant operators, we wanted to ensure that we had also simplified ResyOS in all capacities, making it both powerful and user-friendly. Restaurants have traditionally managed every aspect of their reservation book, so to effectively introduce an element of artificial intelligence to what was previously a completely manual task, we needed to gain their trust. As part of our rollout plan, I created a small restaurant simulation in our Resy cafe to run a usability and algorithmic test on ResyOS Fly. I also mirrored reservation books for a couple of our ResyOS Classic restaurants, and worked the shifts with the maitre’d matching every action they made in ResyOS Fly.

Any new restaurant was onboard onto ResyOS Fly. We started to migrate restaurants from Resy Classic to Resy Fly. Staring with 5% of our restaurants, increasing it to 10% and then to 20%. We maintained tight feedback loops. Once we were confident in the solution, we migrated all Resy customers onto Resy Fly.

Resy Conference: Imagine the Future of Hospitality

Resy hosted its first industry conference in April 2018. Announcing Resy’s Editorial Blog, Resy.com - the web platform for restaurant discovery, and Resy Fly. I was responsible for managing the creative production, the CEO’s keynote address, and the CTOs product demo of ResyFly,