Experience / Service Design

Smart Cart For Businesses


Experience Designer


June 01, 2018


April, 2019


Your business orders online, not on the line.

The Team

I was the individual contributor who worked on this project. I asked for feedback from other designers and managers for feedback to keep improving experience.

The Time

I had five weeks for the design work.

The Deliverables

Two concept testing results, one system map, two usability testing results, different wireframes and mockups, and final prototypes.



Project Abstract

Problem Space: Verizon business orders had to be inquired and placed through phone calls with our phone representatives. It posed several issues to business owners and Verizon:

1. With hundreds of different configurations for business users to select, phone calls were usually very long.
2. It required business owners to write down multiple configurations they would like to try, compared it on paper, asked Verizon to lock pricing for some configurations, and got back to us if they cannot make an instant decision.
3. The whole process was nothing visual at all for business owners.
4. Operational costs were high for Verizon as well.

Solution: I designed this Smart Cart for Small and Medium Businesses as an extension of the Smart Cart for Consumers. In my simple design for the Verizon web channel, business owners can configure up to 12 different plans with auto pricing lock for 7 days, compare them side by side, and lock the price for 14 days manually. All of these can be done on the Verizon business website.

Impact: At that time, in almost all industries, business orders had to be placed through phone calls and it was a well known common practice that no one felt wrong. I challenged this and pioneered a way to allow users to place business orders in the digital channel.

Please contact me for InVision prototype.



Is it now 2018 or 1998?

You must have heard of such narratives: Technology has fundamentally changed every aspect of our lives for the past 20 years in many ways, such as... no, not really so, if you are a small or medium business owner.

If you are a small or medium business owner, and you want to order our fabulous Fios services for your business, you still have to do it an old fashion way: You call us over the phone, talk with a human to figure out some configurations that might fit you, jog down some notes, ask for quotes, and make a decision, usually, not right away.

Just because it is common practice does not make it right. There are enough things for business owners to worry about, so why we still want to give them another headache for something they are paying for? We certainly love our customers, right?

As lucky as I always have been, I got the opportunity to take care of our customers, once again, through my design.


Design Thinking Process

A design thinking process.

I chose the design thinking process for this project because I have enough time to go through the whole cycle of the design thinking process.

A typical design thinking process consists of five iterative phases: Understand, Define, Ideate, Prototype and Test.

Understand: At this stage, you need to understand the problem, users, and their needs, context, limitations, status quo, anything that may be involved or may impact this project in any way. We have multiple ways to strengthen this understanding phase, either an interview, a survey, a lightning talk with the business team, current metrics from the analytics team, or competitor analysis.

Define: At this stage, we need to clearly define the critical information that will for sure impact the direction of the design, which includes but is not limited to the target audience, problem, goals, persona, strategy, and new user journey map or service blueprint, if this is a service design.

Ideate: At this stage, with a deeper understanding of the context and a clear definition of relevant information, we can start to ideate different wireframes, as many as possible. We focus on quantity rather than quality because we want to cover as many good ideas as possible.

Prototype: At this stage, with many different wireframes, we can have some final designs we would like to proceed with. In order to be prepared for the next stage of usability testing, we usually piece them together as an interactive prototype in InVision.

Test: At this stage, our target audience will need to go through one or more prototypes and tell us what they like and what they do not like. Based on the feedback, we can make further improvements.

Although I described the whole process as a linear process, all design process is actually an iterative "Diverge and Converge" process, meaning you would need to go back and forth before you can reach the final design ready for delivery.



Understand the status quo.

I had an unambiguous problem and target audience to start with: Our small and medium business owners want to have the capability to set up and compare different configurations, and place their orders online instead of on the line because it gives them more freedom: freedom of pausing and resuming the ordering experience at any time, freedom of not jogging down some notes in a hurry while on the call, and freedom of sharing different configurations with others.

My initial research partnered with the research team further validated the existence of these pain points. I talked with six small business owners and listened to call recordings shared by the call center, and I identified and validated the following:

1. Phone order calls are usually very long because of complex configurations pertained to business orders, especially when call agents need to explain a specific service or place customers on hold because they need to research something. From the ten recordings I listened to, the shortest one was around 30 minutes and the longest one lasted for almost 70 minutes.

2. Customer interviewees generally do not like the psychological pressure they felt when they were on the calls. Call agents lived on the commission-based model which required them to convince customers to place orders, and upsell them, usually at the wrong time, something customers do not actually need.

3. Due the the complexity of business orders, which contain hundreds of different equipments and services to choose from, small business customers need take time to set up and compare different orders.

4. Four out of six business owners did not have enough knowledge judge if the orders were good fits. Five out of six did not quite trust phone agents because they know agents were commision based. They would like to share what they are about to order with other people to further confirm.

This phone only model not only hurts the experience of our business owners but also drives up the high costs in the call center with apparent reasons. I know, I know, we want to upsell more services or pieces of equipment, and it has higher success rates during a "human to human" conversation, I get it; however, what if we instead offer users a satisfactory order experience, and take advantage of this excellent experience to bring our new acquisition rate to next level?

Statistics from the analytics team offered me another angle to see this problem: The revenue from business orders well exceed the revenue from consumer orders, double the number, and the business market is developing way faster than the consumer market. With so many competitors out there, it is hard for me to justify a conservative approach if we want more shares in this territory.



Define a goal, success metrics and others

It was pretty clear to this project in terms of target audience and goal, with what I shared during the "understand the problem" phase: I want to create a one of a kind digital experience that could offer customers the ability to set up, compare and share multiple orders in the shopping cart, and customers should also be able to lock pricing for further consideration, all on our website.

I then needed to define some metrics to measure success. Due to the nature of this project, I cannot share with you our key metrics, but for all the digital orders we will process, I strived to achieve a satisfaction rate of 72% for online order, comparing the satisfaction rate of 43.9% for phone orders. I defined this goal after numerous discussions with the business team and I was confident that this was realistic.

I also strived to achieve a shopping cart abandonment rate of 70%, lower than what was on the consumer smart cart. I was, to be honest, unsure about this because this was a very aggressive set of metrics, but after I talked with my manager, my manager encouraged me to be aggressive on this number because I was, from his quote, "exploring an uncharted territory."


User satisfaction rate

For users who used this digital experience.


Shopping cart abandonment rate

A lower rate than on the consumer side, a 30-day span.


Opertaional Costs Reduction

in call centers for businesses.

A good service blueprint cannot live without a protagonist, and in the design world, a persona. I created four different personas to represent different sub-groups of my target audience, as well as proposed four different service blueprints. To make the whole design process easy to digest, I only emphasize one persona and one service blueprints here.

And now I asked myself a question: "How Might We resolve the pain points I identified?" Even I was the only designer, I did not want to fight alone because I need "How Might We's, lots of How Might We's" (you know the pun if you are a The Matrix fan). I took the lead and called a group of my designer peers and key stakeholders, and hosted a workshop.

The workshop was fantastic because, besides the pizzas my manager generously offered, every key stakeholder had a chance to express their opinion in the brainstorming session. We identified six categories through affinity diagramming:

1. Ease of use: HOW MIGHT WE make the shopping cart experience easy to use for first time user?
2. Timely reminder: HOW MIGHT WE remind users to come back and complete the order, to reduce the cart abandonment rate?
3. Basic actions (modify, duplicate, delete, save, share, etc): HOW MIGHT WE make basic functions easy to reach while keeping the interface clean and organized through proper visual design?
4. Continuity: HOW MIHGT WE enable users to pick up where they left off when interruption happens?
5. Plain lingo: HOW MIGHT WE reduce the use of jargon and use plain English to describe our equipment and services? 6. Clear indication of status: HOW MIGHT WE indicate configuration status (Complete, incomplete, price locked, price updated, etc)

Inviting key stakeholders to the design process is an effective way to build trust and rapport between the design team and other teams in a collaborative environment because they can feel their voices are heard, it also helps surface the disagreements early and resolve them at the early phases of design and educates other stakeholders to understand the process of design.



"Ideations, lots of ideations." (I love Keanu Reeves!)

I led the team and generated tens of different ideas, merged and picked three of them to further explore the possibilities.

1. Mini cart idea to reduce the complexity of the incomplete shopping cart. It was an idea available in the Apple shopping cart.
2. Two-layer shopping cart idea that shows an overview of all orders and allows users to click further for any order details, with only vertical scrolling.
3. Tabbed design shopping cart idea that exposes details of an order and allows users to switch to another order, with both vertical and horizontal scrolling.

Which one is better? You know who to ask! I created three identical purchase experience with different shopping cart styles, conducted six unmoderated comprison studies, and got the pros and cons for each design from users:

1. Contrary to what I thought, there was no obvious preference between mini cart and full cart.
2. Four out of six participants felt that while the two-layer shopping cart design reduced the overwhelming information about different orders, it was inconvenient when they wanted to switch between different orders.
3. Four out of six participants preferred tabbed design. However, one did express concerns over horizontal scrolling to show additional orders because it may conflict with default mobile browser interaction.
4. Overall, four participants preferred tabbed design while two participants preferred two-layer design.

It really takes time to conduct research, so plan ahead and accordingly to avoid any delays.


Initial Design

Merge pros from different designs.

I analyzed in detail the pros and cons of each design and tried to find a balance in between to generate my initial design about the smart cart. I came up with the initial design below, with an emphasis on important features that could address the pain points I identified during the user research phases. After all, my product cannot deviate from those pain points.


Usability Testing

Usability Testing

Research and usability testing never get old in my design process. With a design that merges most of the pros from the comparison studies I did earlier, I got a surprisingly good testing result in another six usability testing:

1. Six out of six thought this design helped them accomplish basic actions they needed to take during a business order.
2. Three has some concerns over ordering flow, but that was not the focus of this study as I just focused on the experience inside of the shopping cart. They were saved for other discussions related to the ordering experience. I will ignore them here.
3. Although four out of six participants pointed out that inputting a user's email twice to lock the price is not necessary, I still kept it because of the legal reasons.


Final Design

Final Design

Finally, with some minor tweaks, this is the final design I created.

With this design, users are able to create multiple orders, duplicate orders for small changes, delete unwanted orders, compare different orders, lock pricing for 14 days for complete orders, share the whole shopping cart with others, and quick resume with email or phone number.

Due to nature of this project, this prototype will not be available publicly



Impact on both data and emotion.

Data extracted by the analytics team from Adobe Analytics and other tool showed the impact my design made, either on par with the goals I set up, or well exceeded them.

During the managed trial period for over 6000 orders on BAU ordering experience:

1. The shopping cart conversion rate was 4.81%
2. 64.2% of the orders were placed without calling in.
3. Around 29.6% costs were saved in call center.

My design emphasized emotional changes in users as well: On the one hand, it freed users from the psychological burden they often experienced during the long calls with phone agents, allowing them more time to make the right and fit choices for their business; On the other hand, we can even upsell more proper services or devices at the right moment in the ordering flow without inciting "frustration" moments.


Next Steps

Next steps.

I planned to shift the "phone agent" moment, or "live chat agent" moment, to the ordering flow because I only wanted human intervention at the right moments when customers explicitly asked for it. Unfortunately, because of the organizational changes at Verizon, I did not actually take a step further on this.




My Verizon Business Market Smart Cart design was a "one of a kind" design at that moment, and it proved that there were many shackles, commonly known as "common practice", I can possibly break. With Verizon's focus shifting to business markets more and more because it was more profitable, a good user experience for business owners has its own unparalleled value.

I learned two things from this project:

1. Fully understand business requirements by conducting more talking and research. Business-facing design is usually way more complex than consumer-facing design in its functionalities and requirements.
2. Always validate with the target audience whenever there is a decision to make. It keeps me "user-centric".

...I never thought it could be done this way.

Verizon Customer

Research Participant

Simple design with lots of potentials.

Verizon Customer

Research Participant

...this was, my first time seeing this.

Verizon Customer

Research Participant

...overall I love this idea and design!

Verizon Customer

Research Participant

Order online always makes me feel easy...I want to talk with Verizon only when I need to.

Verizon Customer

Research Participant