on-demand messaging experience
Research & Investigate
After a busy week, I realized that it was my mom's birthday that week as well. Since I live far away, I would have to ship a present to her. With limited time, I quickly tried to find a reliable company that could ship a present to her in two days. This is where I found Operator.
What I found interesting about Operator was that it allowed users to request services from a wide range of services through a chat system. Essentially this eliminates the need for users to constantly search for products on sites or in person. Also, it starts a conversation through texting - majority of users that own smartphones are used to texting on their phones. According to Adweek, 49% of of mobile phone users in the U.S. will use mobile messaging apps by the end of 2016. The screens above show the flow users take to talk, select, and purchase an item. Some of the pitfalls I found were
- Adapting to a new interface
- Lack of personal communication between AI system and human and wait time
- Offering different options that meet the request of the user
- Providing payment information initially before beginning experience
- Integration with other systems
I began to try to understand how people shopped online. What were the biggest motivators for shopping online? Do people seek services to help them shop? I researched articles from Pew Research to understand the type of shoppers that would use an e-commerce messaging experience.
Surveys from Pew Research really highlighted an opportunity for an experience to be created. The research really focused on age groups, shopping habits, mobile devices, and lifestyle of a shopper. There are several reasons why users use their phone while shopping: comparing prices, reviews, convenience, and variety of products to name a few. In the research, it's apparent that is an opportunity to combine message with on-demand services. Messenger has become a vital way of communication. Online shopping has revolutionized how we think about services and experiences in the retail industry. After researching, one thing seemed to be clear, requesting on-demand services will be integrated into applications we use every day more frequently.
While researching I developed two personas that helped me focus developing the needs for the experience.
In general, the users are deeply immersed in technology they are on trend with using the latest apps. They constantly use their phone, especially a messaging app. They use services in order to help with their schedule. While users are busy and have limited time, they do interact with people quite often. With having the users in mind when ideating, it raises challenges.
One of the most vital parts of the experience is communication. The experience will focus on users:
Have a normal conversation between the AI + human, like human to humans do
We want to make sure it doesn't appear that you are talking to a robot. It desensitizes the conversation, where the user would not feel respected.
Users should not have to adapt to a new interface
Instead create an experience for users that fit within their habits. The focus is to incorporate the new e-commerce shopping experience through their messenger application on their smart device.
Incorporating on-demand service into day-to-day life
Even though users are deeply immersed in technology, we needed to create an easy and accessible experience for the user to confidently use consistently.
Creating the experience
Below are initial ideas of how the experience would work. I focused on how AI + human interactions could create a normal conversation to the user while delivering on the request.
My goal was to create an experience that didn't require users to learn the ways of another interface on their phone. So I decided to focus on using an interface the smartphone users use on a daily basis - messenger. By starting on their messenger it eliminates the need for users to stop one task for another, finding the product or service is done on the back-end.
The image above explains the process users take to use SOS. The conversation will start by a text or voice message by tapping on the new keyboard, similar to the way Bitmoji works. The main parts of the conversation are - Request, Select, and Confirm.
Within the conversation, certain parts of the conversation switch from AI to human. Specifically, when start the initial request and feedback of the experience is controlled by a bot. Human interactions are focused within finding and select the product for the user. The goal is to create an experience that is natural as possible, especially since texting is the only communication we have with the user.
Below is the ideal flow users will take when accessing SOS.
My sketches focus on the entire experience - downloading app, tutorial, requesting product, selecting product, and confirming product. For this experience, the main area that raised concern was the limited screen size I could work with within smartphone messaging systems since I'm requiring the user to add SOS as a another keyboard.
Initial ideas of how flow from changing keyboard to completing the survey to chatting will look like. Movement at this point isn't completed, I really wanted to focus on the flow of the conversation.
Prototype & Testing
From initial prototyping to the final designs, I realized the difficulty of switching keyboard. By adding an additional click to select the keyboard causes an unnecessary delay. So inspired by the Gboard from Google, I decided to change the location of the keyboard to where the predictive word suggestion lies so users can quickly start their experience. At this point, I was still trying to figure out how to incorporate the voice message aspect to the experience. I looked at various inspiration and how it is currently applied to messaging applications.
When testing some of the feedback I received was: enjoyed having six answer options, including options like "something else" or "other" and a good balance how the conversation moved from keyboard to text message. One important part that needed work was users wanted feedback when selecting choices to know that their choice was confirmed.
After going through testing, I thought it was important to test and get feedback on certain parts of the experience. The video belows show the experience with motions and interactions. Enjoy!