Case Study

Employer Alert Center

Project Overview

The goal of this project was to centralize (and standardize) time-sensitive notifications for Sales and Service Reps. By bringing these real time alerts to the tool where Reps are spending most of their day, we are giving them the best possible chance to make a connection with an Employer* while they are using Indeed’s products. This custom object in Salesforce handles group assignments, but also allows individual Reps to optionally subscribe to additional alerts.

*Employer is how we refer to people who use Indeed’s hiring tools.

 

Company

Indeed.com

Role

Lead UX Designer (2022-Current)
User Research, Information Architecture, Interaction, Prototyping & User Testing

Introduction

Throughout this project, I worked with an amazing Product Team who fully embraced the Product Trio concept. We completed some Discovery up front, but we relied primarily on continuous discovery to incrementally deliver value to the Sales and Service teams at Indeed.

Here are some of the ways this project was successful:

  • aaaa


Problem Statement

  • Our Sales and Service teams don’t know when their clients are actively using Indeed’s products.

  • Teams are being notified after the fact, and are missing opportunities to connect with clients while they are logged in.

  • We believe that we can increase the likelihood of upselling or cross selling, if we can improve the way Sales and Service Reps connect with clients who are actively using their accounts.

  • We believe that we can also improve the Employer experience by providing them with the resources they need to make decisions while they are using our products.*

*The UX team added this last bullet point. Traditionally, our internal Product Teams have not considered the Employer experience because they do not use the internal tools we build. However, we know they are directly affected by our Sales and Service models and should be included in the Service Design.

📜 Additional Context

Indeed originally relied on a homegrown CRM. It was designed for managing a B2C business model. As Indeed grew into the leading job marketplace in the world, the old CRM was not able to support a B2B model, and teams were migrated into Salesforce. However, the new instance of Salesforce had to be designed in a way that it would accommodate the old CRM’s data structure, leaving us with a less than version of the most popular CRM in the world.

The biggest challenges we faced were:

  • Limits to the hierarchical modeling of national and franchise accounts

  • A 72 hour delay in surfacing Indeed product data within Salesforce

Empathize 

🎯 User Feedback

We needed to better understand the goals how the tools and processes being used todayThe following insights were collected via Sales and Service teams, so that we could baseline the current experience:

Sales Teams

  • aaa

Service Teams

  • aaa

Define & Ideate

📝 In order to give Sales and Service teams what they want and need, we needed to . We created this basic set of requirements to help guide our initial ideation:

  1. a

Interaction Model

a

Our Hypothesis 🧐

 

Step 1: Creating Boundaries

Start with an empty floor plan

“Draw” the outer perimeter

Save boundaries with unique identifiers

Add rooms to fill out inner perimeter

 

Step 2: Creating Routes

Add a start and an end point

The SW automatically connects the points via shortest path

Save routes with unique identifiers for CRUD

Repeat as necessary

In this model, users have the flexibility to plot individual points or to combine then and create routes

 

Start with an empty floor plan

Add points individually where users will turn

Specify which points should connect

 

Step 2: Create Routes (Route Mode)

Specify starting point

Specify end point

Repeat!

 

🤝 Compromise

  • Though UX’s recommendation was well-received by internal stakeholders and understood as a viable solution at scale, Eng was concerned that there were too many unknowns

  • We agreed to move forward with Engineering’s model, building onto it to achieve UX’s original goals

Prototype & Test

Goals

  • Document Users' initial reactions to the UI and individual controls

  • Understand Users' mental model for adding spaces, points and paths

  • Observe the discoverability of the help documentation

  • Uncover any usability issues with the wayfinding functionality

  • Understand how Users expect to manage spaces, points and paths moving forward

  • Document Users' understanding of the Lobby Display

Methods

  • Conduct a short interview about each User's experience with room management and/or wayfinding software

  • Perform a short, qualitative usability test

    • Utilize think out loud protocol to uncover insights and feedback

    • Simulate functionality via a combination of design software (Axure) and an interactive prototype

  • Review Lobby Display designs

Iteration

We learned quite a bit from our first round of usability testing and got right back to work. Here are some of the revised workflows:

Adding a space

 

Drawing Routes

 

Adding Points to Routes

 

Editing Routes


 

Visual Design

As this project has been in design, we have simultaneously been updating the Room Agent UI to follow our Design System as closely as possible

  • Our Design System is comprised of 60+ Angular components

  • Room Agent is built on .NET and utilizes Windows controls

What’s Next?

📊 More Usability Testing

🔄 More Iteration

 

Future Problems to Solve

  • Expanding the map view to support a campus environment

  • Explore filtering for spaces based on equipment and other criteria

  • Explore a web view (e.g. students want to be able to view the library Lobby Display from other places on campus so they can reserve a space BEFORE they walk across campus)