Inclusion of Amazon Advisor



In my summer 2023 Amazon internship, I led the design of the Amazon Advisor Inclusion Recommendation project. Our goal was to streamline the addition of Amazon Advisors to the onboarding process by implementing an employee recommendation engine known as the People Recommender pattern.


Jun 2023 -
Sep 2023




TPM ( Sharda)
PM ( Anil )
Mentor (Rebekah)

My role

User flows
Thematic analysis
Functional prototypes
User testing
Design system

What is Amazon Advisor?

As part of the onboarding plan for new hires, managers are asked to designate several contacts to assist new hires in learning about their roles, and an Amazon Advisor (AA) is one of them.

Compared to an onboarding buddy or mentor, an AA provides an outside perspective of your team and serves as the go-to person when a new hire does not feel comfortable talking to teammates.

so, what is the problem?

A low adoption rate of Amazon Advisor
A dissatisfaction from new hires

Currently, only 43 percent of new hires have an AA in their onboarding plans.

A low adoption rate of Amazon Advisor has led to dissatisfaction with new hire's onboarding experience.


Employee recommender engine
Comparison chart between roles
Clearer point of contact instruction

Provide a smart engine that generates automated employee suggestions in the plan creation tool.

Eliminate confusion about the role of Amazon Advisor and get to know this role better.

Provides clear in-product guidance on how to contact an Amazon Advisor and communicate their responsibilities.



300K+ managers were helped
20% increase in adoption rate
Pattern Reusable 10k+ times

Launched after Sep 2023 into the 'Embark' tool, reduced the workload for 300k+managers.

Increased Amazon Advisor adoption rate by 20 % and onboarding satisfaction rate from 60% to 85%.

Integrated 3 variations of design pattern into Amazon design system, estimated to be used 10k+ times.

Timeline & tasks


User persona

To begin, I have set up meetings with PMs to understand the end users of the tool to understand some past frustrations. Our users for this project are managers with incoming new hires.

The onboarding plan creation process

To onboard new hires, managers will add Amazonians to the contact section of the plan, introducing new hires to meetings with Amazonians. This includes Amazon Advisors (AAs). My project involves redesigning this contact section.

First-hand testing experience

The first step for me is to understand the tools and get to know the onboarding plan creation tool better. To do so I have had a first-hand experience of using the tool from a Gamma link to conduct initial pain points on the visuals. Click to see different discoveries.

A confustion between roles

In the current product experience, there is no explanation of what an Amazon Advisor is. As someone who has never used the tool before, I was confused to the differences between Onboarding Buddy and Amazon Advisor.

Unclear Amazon Advisor qualification

Based on the current experience, managers do not have a clear idea to what qualify has a good Amazon Advisor.

In the 2022 Q3 testing conducted by the ONEX group, researchers found that when managers do not clearly communicate the role of AA to selected employees, new hires experience a 15% negative onboarding experience as AA fails to fulfill its responsibilities.

Irreleveant information

On the current contact page, there is a lot of irrelevant information that is unrelated to selecting an employee.

When creating an onboarding plan, managers occasionally omit an AA because they lack familiarity with individuals outside their team whom they consider suitable for the role.

Contextual interview & Journey mapping

Then, I assisted in 4 contextual interview sessions with a UX designer on the entire tool, including the contact section, to discover the pain points and areas for improvement in the manager's onboarding plan experience.

I have then synthesized my observations of managers in the contact section into a journey map to determine the main area of concern.

Framing the big problems

Confusion between stakeholder roles
Lacking proper communication
Limited connections lead to missing AA

Managers often become confused about the differences between Amazon Advisor and the onboarding buddy.

When managers don't clearly explain Amazon Advisor's role, new hires' onboarding satisfaction drops by 15% as AA fails to fulfill its responsibilities.

When making an onboarding plan, managers may skip an Amazon Advisor (AA) because they aren't familiar with suitable individuals outside their team.

business needs and technical constraints

Success metrics

Increase the new hire satisfaction rate from 60% to 80%. Increase the adoption rate of Amazon Advisors (AA) on the onboarding plan from 70% to 84%.

Tech Constraints

According to the tech team, we are currently unable to implement designs that connect Embark with other tools. As of now, only updating in-product guidance and implementing design patterns is allowed.


How might we redesign the onboarding page to help managers add Amazon Advisors to the onboarding plan so that new hires can get involved in Amazon culture more easily?


Brainstorm solutions

Based on the 3 synthesized user pain points, I have brainstormed solutions for all three of the frustrations. I have categorized each of the problems into different colors to identify them more easily.

Feature prioritization

I have then set up a meeting and ask designed on the team to place the three charts on the feature prioritization chart to see which solution offers the best customer value.

main features

Problem #1

Limited connections lead to missing AA


Modifying people recommender patterns, A macro-pattern used to surface recommendations of qualified people, to surface AAs. We hypothesize managers will assign AAs more often and lead new hires to establish stronger connections.

Problem #2

Lacking proper communication


• Providing clearer in-product guidance that reminds managers to communicate with the selected AA

• Banner to reiterate to communicate with AA

Problem #3

Confusion between stakeholder roles


We will be providing a comparison chart, as it clearly and concisely displays the differences between AA and Onboarding Buddy.

The userflow

To further define users’ actions, we created user flows for the redesign to gain a better understanding of how users will use it. Since we are only redesigning a section, the flow is rather linear and simpler.



Qualitative & quantitative testing

Next, I asked 14 managers to complete qualitative testing for the MVP wireframe on UserTesting. In general, we have divided the testing into 4 different tasks to better evaluate how managers behave in each process: Understanding the difference between roles, Selecting an AA, Reselecting an AA, and To-do after the selection of AA.‍‍

I have also included a quantitative survey at the end of each task. The majority of concerns arise in the task of "selecting AA" (owhile "understanding the difference between roles" and "to-do after selection" have a 4 and above rating for ease of use.‍

Thematic analysis

To compile the research results, I employed Thematic Analysis to categorize the qualitative data gathered into various themes and understand managers' common pain points for the redesign.

Positive feedback

1. The chart is very easy to understand.
2. The people recommender is helpful to find employees

Ares need to be improved

1. Hard to choose from so many given employees
2. Redundancy in illustration
3. Confusions on the search box in the pattern
4. Lack of clarity on why candidates are being suggested
5. Where are the colors?


Problems based on feedback

Based on the testing, these are the action items that the managers have pointed out during the session. They include wordings, illustrations, hierarchy, and some redundancy. They also wish to see the addition of colors.

Iterated design

Based on their feedback, I have iterated the parts that need to be adjusted. I have also deleted redundancy and fixed the visual hierarchy. However, there remained one more problem.

Design challenge

We need to modify the people recommender pattern to better align with our specific use case and appear less like a search for recommended employees.


How can we make the search box and recommendation box appear as distinct selection methods?

Brainstorm challenge direction

To tackle this  I have brainstormed 3 different designs to approach this problem.

Option 1: grouped components, search below
Option 2: seperate components, search below
Option 3: side by side

Final decision: side by side

After meeting with the design and tech team, I've chosen the third option:  side by side.

This allows us to prioritize both methods equally without interfering with the "load recommendation" feature.

design system from Stencil

Before handing off the design, I have conducted quality assurance tests with developers and designers to ensure that all colors and components are within our Amazon Stencil design system 2.0.


In addition, I have set up an accessibility review with developers to adjust my final designs to confirm all colors passed WCAG 2.0 AA standards.



Differentiation of the two roles

Provided a simple comparison chart to allow managers to understand their differences at a glance.


People recommender for employee suggestions

Replaced the search box with the People recommender pattern to suggest qualified employees. Having a search box on the side allows managers to compare the recommended employee and search employee.

I have also considered the Error message user case if the manager selected an unqualified candidate.


a more communicative in-product guidance

Prompt managers communicate with AA before their start date with a more personalized start date of the new hire.


Reminder banner

Reminding managers to communicate with AA before their start date after successfully completing the selection of AA. In addition, reminding managers again to reach out to AA to remind them of their responsibility.

Final site map


Key takeaways

Understand design system

During my internship, I facilitated a meeting with the Stencil team to finalize and demo the people recommender pattern. This highlighted the significance of a design system in ensuring consistent cross-platform design. Embracing uncertainty in patterns and considering user journeys across platforms enables the exploration of diverse variations of the same pattern.

Learned to pushback when needed

Throughout the project, I collaborated with PMs and the tech team to outline the business scope and assess technical feasibility. While there might be instances where PMs requested design implementation, it's crucial to consider the customer's perspective and respond with customer-centric approach.