Asurion Sales Portal
COMPANY
ASURION
DATE
07/15/18
DURATION
4 MONTHS
BASED IN
SAN MATEO, CA
MY ROLE
While at Asurion, I led the Asurion Sales Portal team which consists of ten engineers, two product managers, two product designers, and one product design intern. The Asurion Sales Portal (ASP) is the site that carriers sales reps (AT&T, Verizon, Sprint, etc.) use in-store and is Asurion's single point of contact with the sales reps. This was a hands-on role where I was directly responsible for the product's visual design, interaction design, information architecture, and user research.
THE PRODUCT
All of the major phone carriers (AT&T, Verizon, Sprint, etc.) white label Asurion's products and services and sell/distribute them through each of their respective channels. In the higher-priced monthly subscriptions, customers not only receive phone insurance but also AT&T ProTech support which consists of anytime access to Asurion's technical support via its U.S.-based call centers (think of it as Geek Squad in your pocket).
"Onboarding" is a use case generated from a pilot that was first run with AT&T. When a customer buys a new device at AT&T and signs up for device insurance and ProTech Support, the sales rep offers to schedule the customer an appointment so that an Asurion ProTech Expert can assist in setting up his or her new phone via a phone call from the ProTech (Asurion call center agent) later that day.
Onboarding appointment scheduling provides value to the sales reps, the customers, and the business. Sales reps used to spend hours transferring photos, transferring contacts, helping the customers find passwords, etc.- time that they could have spent selling products instead. Onboarding appointments allow the rep to offload these tasks to a ProTech Expert who can help the customer remotely from the convenience of the customer's home.
From a business perspective, Asurion found that this initial interaction on average leads to a significant decrease in churn rate, and an increase in contact rate (a metric that earns Asurion more money from AT&T), which both earns Asurion and each carrier significantly more money, and an increase in WTR (customer satisfaction).
MY CONTRIBUTION
When I started at Asurion, the sales portal where the reps accessed the onboarding scheduler had just emerged from a design sprint and showed early success in a pilot market. During my time at Asurion, I have led all research and design efforts around the sales portal and helped successfully sell it to AT&T, Verizon, and Sprint corporate and indirect stores (an enormous company win). The following projects I feel were key in this success and demonstrate some of the large design and research efforts that went into the product.
1. RE-ARCHITECTURE AND REDESIGN
GOALS
This complete redesign was sparked because of a pending code migration to React, which required rebuilding the entire site. It was because of this, that I felt it was the ideal time to redesign and re-architect the site in parallel with the react migration to accomplish the following objectives:
​
Primary Goals:
-
Decrease time to schedule an appointment by 10%
-
Increase dashboard views by 50%
-
Increase content views by 20%
Secondary Goals:
-
Refresh the site's looks and feel to offer a cleaner visual aesthetic.
-
Improve information architecture to ensure that the new designs support the mental model of the sales reps and managers.
-
Establish a pattern library that will allow the team to scale to new carriers with minimal dev/design resources
-
Provide a structure that will allow for expanded training content in the future.
-
Design a structure that will allow for more tools/functions in the future.
Screenshot of the product when I started on the team.
RESEARCH & DISCOVERY
I conducted several forms of information architecture research that led to a successful restructuring of the site's content. The steps I took are as follows:
​
-
Open card sort - Recruited AT&T store reps/managers and Asurion RAMs (outside reps). The goal was to discover logical groupings and generate names for these groups of content/features. Because the product was still in its MVP state, I conducted this card sort by not only asking users to sort current features but also potential features in the future after meeting with the product management lead.
-
Closed card sort - The goal here was to validate the groupings and names that came out of the open card sort.
-
Tree test - Helped ensure that the structure and names even without any styling made sense.
-
Conducted usability testing with new concepts at point of sale in a test market.
​
In addition to the IA research, I conducted extensive usability testing around the current interface to identify problematic opportunities.
The dev lead on my team (right) helping to administer an open card sort with an Asurion RAM.
FINDINGS
IA research takeaways
​
-
Content needs its own section
-
Unique grouping of metrics
-
Separate tools section
-
Global scheduling feature
​
Usability testing takeaways
​
Content library:
​
-
Users did not realize there was more content behind the menu that was used previously.
-
Not only were there no signifiers around the previous point, but there was no way for the user to tell which might be an article or a video (I learned that reps have drastically different learning styles making this point important).
-
The structure for the content before did not create a pattern that would allow for the addition of any more content (it wasn't future-proof).
-
Users perceived the menu design previously used as a component that would open content on a new page based on iOS standards.
​
Scheduling flow:
​
-
There were severe usability issues with many of the components within the scheduling flow that were not optimized for a touch-friendly interface. AT&T reps visit this portal on their iPads so it was vital that all components be touch-friendly.
-
Users were completely missing key fields that weren't required, causing issues when appointments were being booked for incorrect times.
​
Dashboard:
​
-
By default, the dashboard was not optimized for the information that reps were looking for.
-
Reps did not notice the segmented controls that allowed them to switch date ranges
-
By default, the dashboard showed a weekly view, but reps indicate they think primarily about monthly stats
-
By default, the dashboard showed the store-level stats which proved to be only important to managers. Reps primarily are concerned with their own stats.
-
Reps did not realize that there was additional content available beyond the initial video that is viewable from the dashboard.
​
Data and Analytics
​
-
The primary problem revealed in analyzing site traffic was the limited number of page views on the dashboard. When the rep accessed the portal, they were dropped straight into the scheduling flow and would end the flow on an appointment success screen. From this success screen, there was very little click-through to the dashboard, which was a severe problem for the team.
VISUAL DESIGN AND USABILITY TESTING
To address the usability issues with the scheduling components, I completely redesigned many components, allowing for a more touch-friendly interface. These improvements led to an 18% decrease in time to schedule an appointment.
As mentioned above, when viewing the analytics, there was little traffic going to the dashboard and reps were not discovering other parts of the portal. To address this issue, I removed the full-screen success page and instead moved to a pattern that drops the user on the dashboard after an appointment has been scheduled and shows a success banner that can be closed.
​
This change increased dashboard views by over 1,000%. This was a small but crucial change that helped to increase rep awareness and engagement with the rest of the portal.
For the content, I chose to break it out of the dashboard and create a separate section for it altogether. This helped to solve some of the usability issues detailed above, and also helps to create a foundation to further gamify the experience in the future. This change resulted in a 42% increase in the number of content items viewed per session.
From a navigation perspective, I decided to ditch the top-level tabs and instead opt for side navigation. There were several reasons for this change based on the IA research:
​
-
Reps understood the scheduling tool to be a global feature so I created a full-screen modal pattern that can be accessed anywhere in the application.
-
Reps expected the training content (articles/videos) to be outside of the dashboard. Because of this, I broke up the sales training into a separate tab altogether.
-
Product management hoped to grow the suite of tools in the future in accordance with rep needs. This pattern allowed for that expansion without nesting tools out of sight.
​
In addition, as seen here on the dashboard I reset the default views to show a monthly view, as well as individual stats for the reps and store stats for the manager. A handful of other usability improvements were included as well, such as the removal of the page-level date selector which was not immediately discoverable to reps.
PATTERN LIBRARY
One of the major pieces in the redesign was creating a component-based pattern library. I modeled the organization using atomic design principles of atoms, molecules, etc. Initially, this library was built in Sketch as shown, but we have since moved it to Figma. We successfully implemented this while keeping the style variables to a minimum to increase the speed of launching new features between carriers, enabling our team to move very quickly.
2. NEW DASHBOARD AND LEADERBOARD
RESEARCH & DISCOVERY
While performing discovery research in Chicago, there was a characteristic that emerged which was consistent between both managers and store reps; competitiveness.
​
At the store rep level, we found that reps were extremely competitive with each other. One of the ways that this was on display was in their group chats. Each store is part of a group chat messaging app, and within this app, we found constant examples of playful "trash-talking" between reps about personal performance within the store. Another example was found in their Samsung PRO app. Within this app, there was a gamified experience that rewarded the user based on their knowledge of Samsung products.
​
At the store manager level, we found a great sense of competition between stores. Managers were not concerned with their own personal stats, but when it came to store stats, managers felt a great sense of ownership and felt compelled to do better than other stores in the area.
The gamified Samsung PRO app effectively helped to motivate sales reps to learn more about their products.
Initial designs and usability testing
Based on the key learnings above, I worked to validate an MVP of a leaderboard to help motivate store employees based on their innate sense of competition. We choose to use a two-tabbed leaderboard and allow users to toggle between store stats and personal stats.
​
By default, sales reps view the rep leaderboard to play on their rep-to-rep sense of competition and managers view the store leaderboard by default to play on their innate store-to-store sense of competition.
​
Process-wise, I began by producing low-fidelity designs and tested the UI in-store to uncover any usability issues before launching the new designs. I recruited another product manager and we thoroughly tested these designs in Detroit (we were limited to conducting research in specific markets at the time).
IMPACT
The impact of the dashboard redesign and implementation of the leaderboard feature has been incredible. This feature has driven an increase of 25% in active sales reps all while increasing the average number of appointments booked per rep by 22%.
​
While following up in-store, to gain more context behind the analytics, it became clear that our assertion about their competitiveness was correct. We would consistently see brand new reps (high turnover in carrier stores) who didn't have a clear understanding of the platform as a whole, but they almost always know exactly where they are on the leaderboard.
​
"I'm not sure if I've seen the dashboard before. But I know I'm #1 on the leaderboard."
​
-ATT Sales Rep
3. APPOINTMENT HISTORY FEATURE
RESEARCH & DISCOVERY
Through user research, a need surfaced to “close the feedback loop” with the rep. Reps continually asked, “How do I see whether an appointment was kept or not.” In addition, closing the feedback loop was requested by store managers as a way to increase rep engagement with the sales portal over time. After presenting and iterating upon a high-fidelity prototype, we felt very excited about the early feedback we received in-store and decided to launch it.
DESIGN AND FUNCTIONALITY
The design and functionality are meant to be simple and straightforward. Each card displays basic information about the appointment as well as whether the appointment was kept or not by the customer. One of the main reasons that the rep likes this functionality is so that they know whether to follow up with a customer or not to ensure that he or she receives all the help that they need. This contributes to their WTR score (customer satisfaction).
IMPACT
The primary impact of this feature ended up being an increase in the rep's WTR score. For reference, the WTR score is a customer satisfaction rating from the customer that each rep is compensated on. Over the course of six weeks after launching the feature, we saw the average WTR score of each rep increase by 8%.
​
But why?
​
The increase can be attributed to the rep's ability to follow through and ensure that the customer had a good experience. Sometimes the customer is not able to make their appointment and for whatever reason, misses the phone call and this feature empowers the rep to go the extra mile in that scenario and ensure the customer has resolved all of their issues with their phone.
​
One of the other unanticipated benefits of this feature has been the insight into a feature that is currently in development; the rescheduling feature. After launching the appointment history section, we started getting requests via Intercom from this page in particular with the request to reschedule appointments for the customer. The customer already has the option to reschedule appointments themselves via SMS, but it was clear the non-tech-savvy users relied on the rep to do it for them, and reps expected the functionality to live here within the appointment history section.
4. NEW APPOINTMENT TYPES
RESEARCH & DISCOVERY
While doing in-store discovery research, I heard from store reps that appointments were being booked for reasons other than new device setup. I was able to validate this by speaking with additional reps, it was estimated that between 30% of onboarding appointments are actually being booked to help a customer with his/her existing device.
Shortly afterward, I went on a research trip to Phoenix to shadow our call center agents. We heard directly from agents that between 30-40% of calls they take are also for an existing phone, and not a new device setup. One agent also mentioned they he knew when a call would be about an existing device if he sees that the rep entered an identical device for new and old devices within the scheduling flow. Upon analyzing this data in our database, we were able to validate this quantitatively as well.
​
Up to this point, appointment scheduling was designed for one purpose; to help get the customer set up when he or she purchased a new device. From a UX perspective, this meant that we were not providing the ideal scheduling for for those looking to book appointments outside of new device setup.
DESIGN & FUNCTIONALITY
We came up with a solution that takes into account the feedback above by introducing new appointment types. The initial appointment type for New Device Setup is still present, however, there is now a new "custom" appointment type, as well as additional appointment types for specific reasons.
​
By proceeding with a "Custom" appointment, the user has a slightly shorter flow and is highly encouraged to leave specific instructions based on the large visual prominence of the text box within the flow.
​
The additional appointment types were created by analyzing the data that has been entered by reps in the past when we were able to determine that the appointment was not for a custom appointment. We then set in place a cadence to check what information is being entered for "custom appointment" after launch and create specific appointment types for the most popular trends that emerge.
​
​
IMPACT
Immediately after launching, we found that our initial research proved accurate. 29% of all appointments booked can be attributed to "custom appointments" and an additional 10% to the other various appointment types.
​
This was an instrumental win for the team and Asurion because it proves that customers are relying heavily upon sales reps to solve their technical issues, and those reps are relying upon Asurion to successfully save them time and money.