Amazon Business Customer Care
Overview
Building a customer service experience for Amazon’s
B2B customers. Designing the primary customer support tool to serve Business customers, driving user research and ideation efforts.
Role
UX Design Lead
Team
Amazon Business Post Purchase
Status
Incubated in US marketplace Dec 2021
WW roll-out in progress
Background
Amazon Business Customer Support is a multi-channel, multi-tier support ecosystem that caters to customer queries across channels such as phone, email and chat. The Customer Contact Handling tool is the primary system used by Amazon Business Customer Support Agents to drive resolutions to customer issues.
My role
I lead the UX for the primary customer contact handling tool for Amazon Business. As part of my work, I collaborate closely with Product, Research and Tech to analysis customer contacts, identify and prioritize customer & associate pain points, craft user stories, design, test & validate UX flows and deliver final designs for implementation.
Some of my key contributions include:
• Implemented new order management flow. I designed and implemented a new order management flow to enable customer service associates to manage customer contacts for order-related issues.
• Enabled business grade experience across the tool. I identified critical areas within the contact handling tool and enabled business grade features to enable associates to serve multi-user business accounts as well as single-user business customers with ease.
• Led user research & validation. I led UX Research efforts by planning, conducting & synthesizing user studies with internal stakeholders. I developed repeatable mechanisms to gather user insights & feedback at each stage of the design process to ensure solutions focused on the right pain points.
• Established design patterns and UX paradigm. I developed new interactions and design assets that formed part of the design library for the tool along with defining the core UX paradigm that enables other designers to develop new features & flows for the tool.
Understanding the problem
Amazon Customer Support is the primary customer service tool aimed at handling customer contacts including call, email and chat support. With the need to cater to increasingly complex business queries, there is a need to develop a solution that meets the needs of Amazon Business.
The current experience is an information heavy model and is highly CSA driven. Research indicates that the current model has a high learning curve resulting in increased ACHT, low accuracy of SIC matching and low RAP.
65% of contact issues are not accurately identified, leading to high service cost and loss of revenue.
Average call handling time (ACHT) increases for more complex customer queries involving multiple screens.
User insights
Amazon Business Customer Support comprises 1000+ CSAs at 12 CS locations supporting 9 countries (DE, ES, FR, IN, IT, JP, UK, US, CA) and 8 languages (English, Spanish, French, German, Italian, Hindi, Japanese & Chinese).
Amazon Business Customer Service has a tiered structure to cater to Business issues and escalations that includes Generalists, Enterprise Associates and Customer Advisors etc. Researching into common pain points revealed key insights on how Associates use the current tool.
Users find it
difficult to
Understand customer roles and permissions
Identify the correct resolution paths for customer issues
Process multiple resolutions for a group of orders
Users basically
resort to
Utilize browser search to identify user & order information
Select issue code based on intended resolution path
Repeat steps to process multiple resolutions
UX Deep dive
As a first step, I connected with a group of tenured Business CSAs to go over critical use cases over multiple 90-120min sessions. The goal was to understand the task breakdown from issue identification to resolution mapping to contact wrap-up.
Key findings
- 👨🏻💼🧑🏽💼Business customers typically have a complex organizational structure of buyers, procurement managers, finance managers, admins etc. who utilize various features of their Business Account.
- 🚚 📦 Business Customer contacts deal with complex scenarios involving large orders, groups and users, payment & invoicing etc. The consumer contact handling experience is built for retail needs and does not cater to the complexities of Amazon Business use cases.
- ☎️ Amazon Business has a multi-tier support structure that requires CSAs to have more control and flexibility over the contact handling experience.
- 🧑🏼🔧 There is a high dependency on recall vs recognition for core features and workflows. This means that tenured Associates generally perform better than those with less experience.
Understanding the retail experience
I connected with the consumer contact handling experience team to understand the UX framework defined for Retail customers. The consumer CS experience utilizes a simplified framework that focuses on progressive disclosure, linear workflows and an emphasis on customer data security.
Since the solution for Business customers was based on the retail framework, I worked on identifying experience gaps that would need to be solved to provide a robust support experience to Amazon Business customers.
Utilizing the retail UX framework, I conducted a user study with a focus group of Amazon Business CSAs to identify pain points from a business perspective.
Key findings
Business Accounts have unique roles & permissions that need to be surfaced to the CSA to take key actions.
‘..Because in Business account, administrator has certain rights. On the very first page, we would be able to understand what permissions have been given and who has the ability to take what sort of action on that account..’
Business Orders have larger quantities and requisition history that need to be surfaced to the CSA.
‘..if customer has asked about Asus Lite Portable USB, it would be very difficult for me to find it on this screen [CLP]. I would have have to click here [Items hyperlink indicated] Surely we can save some seconds here and save on CHT for CSAs..’
Business order identification and resolution requires additional meta data to be surfaced to the CSA.
‘..sometimes in call, customer may not give order ID, customer only gives name of the product which they have ordered. In this case, I cannot go and ask for order id because it does not look good. We are here to ensure that we are minimizing the customer’s effort…’
Defining the North Star
Based on the research and user connects, I formulated a product vision for the new Amazon Business Customer Support contact handling experience.
Enable Customer Support Associates to anticipate customer problems and provide relevant resolutions with minimum effort and time spent talking to customers.
Product vision
Product Tenets
Relevant
Right information at the right time to the right user.
Flexible
Flexibility and control to both pro users as well as novice.
Frictionless
Frictionless experience to end customers.
Defining experience goals
I conducted research on the best practices in Customer Service management. This helped define the vision for Business grade customer support and enable design decision making.
Speed
Responsiveness has the highest impact on both customer satisfaction (fast response) and dissatisfaction (slow response).
Accuracy
Customers regard accuracy as the minimum and inaccuracy in information & resolution offered leads to higher dissatisfaction scores.
Clarity
Clarity in communication drives a positive customer experience. When accuracy is the what , clarity is the how.
Transparency
Customers value being walked through the actions taken by the associate even if it makes the contact time longer.
Accessibility
Research indicates that reducing customer effort leads to higher satisfaction scores.dissatisfaction scores.
Efficiency
Efficiency is a crucial factor in customer service leading to lower Average Call Handling Time, higher First Response Rate and First Contact Resolution Rate.
Prioritizing UX
As an Amazon Business Account offers business features such as quantity discounts, invoicing & payment solutions and multi-user account settings, there was a need to map business centric use cases that drive a high volume of customer contacts. In order to ensure the team prioritized the development of most relevant features, I partnered with Product and CS teams to identify critical business use cases.
Designing the Large Orders experience
I focused on large order scenarios with bulk quantities for Amazon Business which comprises of 0.5% of all Business orders but makes up to 26% of the order value for key customers. I started with developing the main user flow for solving informational customer queries. Some key pain points to solve included:
Single item may have a large number of packages it is shipped in.
Currently, all packages for a single item are surfaced together in the item solve card making it difficult for the CSA to navigate through the solve card to take any action.
Packages cannot be searched/ filtered through to find a specific package
CSAs may need to identify and take action on a specific package. In the current experience, there is no way for the CSA to search or filter through the list of packages using key identifiers such as delivery status, delivery date etc.
Single item large purchases may be split into multiple orders resulting in multiple cards being surfaced for the same item
The current framework renders cards for each ASIN in an order. In case of large quantity orders, the purchase may be split into multiple orders and shipments in the system. There is no way for CSA to view this information and take any action resulting in repeat actions by CSA on multiple orders.
User flow
I started with developing the main user flow for solving informational customer queries. I also designed and iterated on the concessions workflow that enables Associates to process customer requests such as Returns, Replacements or Refunds.
Design explorations
I explored various paths for high value actions such as returns, pick-up address selection, refund processing etc. I tested design explorations against different use cases including informational and action-oriented. This helped refine the flows and develop reusable design patterns & interactions.
User testing & validation
Since the primary users of the tool were internal stakeholders, I worked closely with the CS org to develop a mechanism of user testing and feedback with a core focus group of tenured CSAs. I developed high fidelity prototypes and tested key scenarios across 2 groups of CSAs on the overall effectiveness of the flow.
Key findings
• ✅ Users found the consolidated packages view easy to use and informative.
• ✅ Users were able to understand & execute the return processing steps.
• ✅ Users found the option to apply common resolutions to multiple items very valuable in reducing contact handling times as well as repeat contacts.
Outcome & Next steps
The new Amazon Business Customer Care experience has been launched WW in 2022. With this launch, Associates have been able to solve over 43% of all customer contacts with a significant improvement in average call handling times.
- Projected reduction of avg. call handling time by XX% WoW
- Projected improved Repeat Adjusted Response Rate by +X.XX% by 2023.
For confidentiality reasons I have omitted the actual values for these metrics.
As next steps, I am developing additional business specific use cases including Tax & Promotions that enable associates to deliver enterprise grade customer support to our customers.
Nithya Asokan Portfolio 