Matthew Paul
Hero image for this case study

OpenTable Automated Guest Tags

Senior Product Designer

2021

In late 2021, I worked with OpenTable as an independent contractor to design a new tool for restaurant owners and operators that allows them to create and manage automated guest tags. These tags can be configured by any kind of business rules and criteria that are then automatically added to a restaurant guest’s profile once the criteria are met by that guest.

For example, “If a guest orders more than one bottle of wine, each over $60.00 USD, on four consecutive visits, add the ‘Wine Connoisseur’ tag”, which the front-of-house staff will see in the guest’s profile on their next visit; along with visit notes to instruct the waiter or sommelier to bring the loyal patron their favorite bottle of wine free of charge.


Product Preview

Effortlessly create automated tags

Choose a pre-defined tag type, configure and customize the criteria based on whatever business rules you’d like, and assign the new automated tag to all your restaurants or specific locations.

Get a birds-eye view of all your restaurants and tags

Filter to any individual restaurant or restaurant group in the Client Admin section of the platform, and manage all of the automated tags that you’ve configured.

Edit or delete tags in one place

Using data, you can determine if you want to keep, edit, or delete any automated tags—and the deals or extra attention to service that comes with them.


What wasn’t working?

The Problem

Restaurant owners and operators of some of the best restaurants in the world were finding it difficult to provide an exceptional and personalized guest experience—that they’re known for with high-profile guests, or customers they remember—to all of their patrons. They wanted a way to capture intel on guests’ visits, spending habits, and preferences so they could more easily act on that data to deliver a world-class, unprecedented, dining experience to every guest that walks through the door.

The exising IA and Navigation into the automated intel tool before we started the project

Discovery & Insights

Through speaking directly with restaurant owners, general managers, and internal subject matter experts (SMEs), I uncovered a few important details that drove the strategy of this product offering.

  1. These customers manually tagged each individual guest profile when they had the time, trying to memorize guests’ preferences; at scale, this was wasting hundreds of hours of their time per year.
  2. Different tag types could be applied to a guest—visit tags for the reservation card and guest tags for the guest profile; OpenTable had a legacy tool to view, manage, and add tags manually.
  3. OpenTable had acquired this tool called Venga Automated Intel, which attempted to solve the challenge, but felt more like interacting with a backend database with duplicate ways of doing the same tasks, pitfalls that caused you to have to start over, and pre-determined categories followed by a seemingly random selection of 60 checkboxes per category (of which only 2 or 3 were usually used); furthermore, there was only a single person at OpenTable who knew how to do this setup for restaurants—it was a white-glove, one-off service.

As-is journey mapping

To complete my understanding of the problem space, and to foster a shared understanding with the PM and other stakeholders involved, I led a collaborative workshop in which the SME & I mapped out the 5 as-is user journies to identify pitfalls, pain points, and begin to note opportunities for an improved experience.

These user journeys included:

  • Navigating to the manual intel configuration tool
  • Editing guest tags for only a single restaurant ID (RID) at a time and undoing changes if there was a mistake made
  • Adding a new guest tag for only a single RID at a time
  • Deleting guest tags
  • Adding a new guest tag without selecting or editing a RID

Opporunities for improvement and innovation

Our hypothesis

If restaurant owners and operators can configure and manage the business rules and parameters for guest tags on their own, for a single restaurant, or across their entire group of restaurants, these custom guest tags will automatically generate when the criteria are met and populate in the Guests’ Profiles for hosts, reservationists, and servers.

These Automated Guest Tags will allow front-of-house employees to provide unprecedented, data-driven, personalized experiences for all of their guests—leading to more loyal restaurant partners, less churn to OpenTable competitors, and furthermore reducing time, manual labor, and money spent on providing the current one-off setup every time for select restaurants.

To-be user story mapping

User story mapping is a helpful tool when ideating, answering technical questions, defining product requirements that are “must haves” vs. “nice to have”, and even begin thinking about prioritization and multiple project milestones.

For this project, I led this collaborative session with the PM and Engineers on the team to even better solidify our shared understanding thus far and to bring them along in the design process.

Define what success looks like

It‘s important to clearly define what success should look like so that the entire team and stakeholders have quantitative or qualitative metrics to measure against, and are aligned on meeting those goals. These success metrics can be used to make decisions in team discussions, for iterating and testing multiple different ideas, and for tracking impact post-release.

  1. 30% of restaurants create at least one Automated Guest Tag within 3 months
  2. 0% churn attributed to lack of Automated Guest Tags
  3. 50% of Guest Profiles include Automated Guest Tags after 6 months of a restaurant configuring them
  4. Owners and operators spend less than 30 minutes per week creating and managing Guest Tags

Initial explorations

IA and navigation

In the pre-existing information architecture and navigation to the legacy, manual Venga Intel Configuration tool, there was a second top navigation bar with Select elements that didn’t match the design system.

The exising IA and Navigation into the automated intel tool before we started the project

One option I explored was to eliminate this second top navigation bar altogether, bringing the “Relationship Management” title and Select components down into the sidebar, and also consolidating from three Select menus to one.

One option for improving the IA and Navigation that I explored

Then through some more iteration, and in collaboration with other designers familiar with more of the surface areas in Relationship Management, we landed on the best option of rethinking and rearranging the sections of this sidebar navigation—giving Automated Guest Tags its own home.

A second, much better option for improving the IA and Navigation that I explored. This is the design solution we ended up building for this small part of the project.

Prototyping

After figuring out the IA and navigation solution and aligning with engineering so they could get started building, I began mocking up and prototyping some early sketches of the primary user stories—or jobs to be done—that we needed to build.

First prototype vignette of a possible tag creation flow

In this scene, the user has pre-selected multiple restaurants and enters the flow of configuring an automated tag for them.

I put this in front of our internal SMEs and learned a few insights:

  • Restaurant IDs don’t matter at the top-level table view
  • The UI was promoting tag creation for individual restaurants too heavily with primary buttons in every row
  • Selecting restaurants as the first step was too limiting, and would cause the user to start all over again if they needed to make a change later

Iterating based on feedback from the first round of prototypes

This iteration shows a more complete journey in which the user can edit existing tags, and create a tag for an individual restaurant or multiple restaurants. I also moved the selection of multiple restaurants into the configuring flow.

At this point, I was ready to put this in front of some customers. Here’s what we learned through some rapid usability research:

  • Editing the configuration of tags is rare, so while there needed to be a way to edit, it didn’t need so much attention at the top level.
  • The first step in the configuration flow felt a little bland and lacked guidance to help the user move through the form fields.
  • Out of the 7 pre-defined Tag Types (that we launched with for MVP), users didn’t know what they were or when to use which one—so I added an extra step in the configuration flow and brought in brand elements and language to help educate the users without compromising agility.
  • In step 2, the majority use case is to have all restaurants selected by default; and overall there was too much redundancy of information across the whole flow.

What we launched

After a few more smaller iterations, we landed on a solution for filtering the table of restaurants, creating and configuring the automated guest tags, and being able to manage these tags by either editing or deleting them with a more subtle interaction.



Made in Colorado

©