Hiya Helper
Hiya Helper is an AI-powered assistant that makes voice calls more efficient and joyful.
Just like a real personal assistant, Hiya Helper can make tedious calls on your behalf. It can also:
predicts hold times
offers both voice and text modes
transcribes conversations
syncs information from calls with other apps like your calendar or wallet.
January 2024 - August 2024
My Role: UX Design Researcher
Other Teammates: 1 UX/UI Designer, 2 UX Researcher
What is Hiya Helper? Is Hiya Helper cool? Is Hiya Helper useful?
What is Hiya Helper? Is Hiya Helper cool? Is Hiya Helper useful?
My involvements
(My brain children)
-
Guide the team with the Double Diamond framework to accomplish a project with minimal existing work and a desired high fidelity prototype.
Lay out Jobs-To-Be-Done (JTBD) framework to clarify the prototyping objectives between research findings and the design process (in between the two diamonds)
-
The initial ask was “develop a product to reduce spam call”. That’s it, there are many divergent pathways to take since there are no constraints.
Therefore, I took lead to tackle the initial research by breaking it into curiosity zones to have sub secondary research so that the team can better define further research directions
-
Collaborating with other UXR, I have produced the necessary materials for the generative user interviews, such as discussion guide and note-taking sheets.
Participated in 4 user interview session being either the moderator or the observer (note-taker)
-
I have coded and analyzed the user interviews that I observed. Progressed from proto-finding, to coded findings with themes, to actionable insights to kick off product design cycles
-
The product direction was pivoted before the design process started.
I proposed this pivot based on the findings uncovered during researches, which shows the original direction was not the root problem and mismatched with user’s need
-
As the only team member who can do research and design, I have been supporting the lead UX/UI designer to complete design tasks.
During this phase, I have designed the full user flow with wireframe screens, information architecture, and other higher fidelity prototypes in Figma, directly partnered with the designer
Problem Space:
Expand the Hiya Suite of app
Hiya’s primary mission is to secure voice with trust, identity, and intelligence.
The Ask:
Explore new opportunity to enhance call labelling and voice security —> Less spam call, reduce risks of Audio Deepfake
Secondary Research
By breaking problem into smaller steps
Since the initial problem space is very broad, I proposed the team to go “breadth over depth” to explore opportunities within different industry that heavily utilize voice call. Although Hiya’s current client are big telecommunication corporation, I suggested to drop the company size constraints and research by business area to look for common themes among different business area.
I have also asked everyone to pick one competitor and quickly note down some strengths and gaps of their top-selling feature.
As it is still early on in the process, we have three researchers and on-time in our plan, it is better to gather more inputs for future reference.
From research explorations, we found that regardless of the industry, users are using personal phones for business purposes to make and receive calls.
-
Smaller Health Care Provider
93% of people say that they want calls from their healthcare providers, but 76% say that they’ve ignored or declined a call from their provider because they didn’t know who was calling.
-
Recruiter and Career Development
Recruiters have high call volumes, and are unable to get through to potential candidates because of lack of caller ID and/or is automatically blocked.
-
Restaurants and Food Delivery Service
Phone calls still play a significant role in facilitating various transactions; for example, of those who have made restaurant reservations, 72% said they usually make their reservations over the phone.
-
Ride-Share Drivers
Drivers frequently have a need to contact their passengers outside of the ride sharing app’s window of contact to coordinate the return of forgotten items.
Define Research Scope
With secondary research findings, the team decide to focus on BYOB (Bring Your Own Device) folks. Still, there are still a lot of ideation and domain definition before the research starts.
My Responsibility
During this phase, I led the team to
narrow down research domains
define research questions with appropriate mixed methods
create qualitative research materials, such as discussion guide and note-taking forms.
The goal is to lead and facilitate conversation to diversify potential research pathways, but also to avoid “too many chefs in the kitchen” problem without making meaningful progress.
Therefore, I promoted “dot voting” process where everyone writes out their idea, explain, and vote among the team. Sufficient discussion is always needed before compromises.
Dot voting to finalize sub research questions
Initial ideation questions about stakeholders and their phone use cases. Discuss as a team to facilitate research question generation
Initial research questions, and method ideation to blend in quantitative method for qualitative result validation
Converting all research questions into potential interview questions, with voting on the best options
Research Questions 💭
1. Why do people use their personal mobile devices for work
What apps are installed and used for work?
What is the purpose of these apps?
What are their voice use cases?
What pain points do they experience within these voice use cases?
Method 𝌣
Generative User Research
What: Semi-structured interviews
Each scheduled session is assigned with 1 moderator and 1 note-taker among the four of us.
My Role: I moderated 3 sessions and note-taken 2 sessions.
Participant 🙆♂️
Group 1 - individuals at Large Companies (n=5)
Participants who use their personal iOS/Android mobile devices for work
Company Size > 5000 people
Full-time employed
Analysis 🤔
2. What are their concerns (if any) around voice security?
How have you thought about solving them (if at all)?
Do the different participant groups (i.e. self-employed individuals versus individuals at large companies) think about voice security in different ways? If yes, how?
Why: Semi-structured interviews maintain a basic framework, but also provide flexibility to allow for deeper exploration of topics. This method’s adaptability makes it ideal for exploratory research and gaining comprehensive insights.
We recruited 10 total participants, equally split between two user groups based on the business size that they are working for. Referring back to the research questions, this user group split is to understand the how employer is influencing user’s phone usage and voice call concerns.
Group 2: Self-Employed Individuals or Contractors (n=5)
Participants who use their personal iOS/Android mobile devices for work
Company Size = 1 OR 2-20 employees OR 11-50 employees
Small Business Owner = Yes
Seniority = Owner
Worked for at least 6 months (i.e. not retired, not recent grad, not unemployed)
I have analyzed 4 out of 10 interviews and constantly synced with the other researchers about their progress. The analysis part is a team effort because the strategy is to extrapolate and consolidate bigger themes into more concrete findings, recursively.
Since we have larger two research questions with multiple sub-questions, this method will help the team to capture finer details of the research findings and attempt to answer the sub questions with more contexts to the larger research questions.
Convert participant's quote into sticky notes and place them under themes
Split off the themes into specific contexts and re-organize the participant quotes
Summarize the quotes under categories into individual sticky notes, and place them into more specific categories
Write out proto-finding after rounds of refinement. Ready to discuss as a group to generate formal findings and insights
Product Pivot from Findings
research direction ≠ product direction
Most important finding:
Hiya’s original messaging of “voice security” did not resonate with participants, unless we specifically provided detailed explanations of the term and examples of voice security being compromised.
Still, after understanding the risk, participants were generally not concerned as they viewed security lapses as “inevitable”
This finding signaled that we misaligned the project direction with user’s actual interest :(
The pivot:
Shift the product direction to enhance productivity and efficiency for phone calls.
A confirmed user interest area, based on existing research result from Hiya
Inspire from existing design assets and concepts of an internal Hiya product.
“I don't care. I'm not worried about it. I know that. I just don't. It's not something I worry about. It really isn't, and maybe I should. But I don't. I just cannot have that worry, because I know that there are big threats out there. But if they're coming... I mean, it's gonna happen.”
— Participant 5 (Large Company)
“It's seems like a necessary evil. You need to use the Internet. And you need these apps and these programs to make your life easier. And then you kind of have to deal with the consequences of stuff like that.”
— Participant 4 (Self-Employed)
“I know that they’re noticing what we're saying, they're recording what we're saying... It's just the cost of having the phone.”
— Participant 9 (Self-Employed)
Design Facilitation
The new product is going to be an AI Assistant that handles phone call with robot call menu and waiting on hold for agents. The assistant has its own app, where user can make request and check call status. We named it Hiya Helper.
Information Architecture (IA)
I have designed the IA based on user action, frontend, and backend to show the feature of Hiya Helper and the logistics of basic user navigation. In the early stage of design, IA also helps the designer to understand how many screens to prototype and how would different screens be connected to each other by logic.
Wireframe
In addition to the IA, I have also designed some mock-up wireframes to showcase some examples of IA items. These wireframes are low fidelity and the design is directly adopted from Hiya’s previous assets that are not relevant to Hiya Helper. Since the design and aesthetic does not matter, it is useful to guide the designer on key interactions and screen elements of the desired interface, and this can be served as inspiration for future iterations
Storyboard
As a team, we made a storyboard to showcase the promised features of Hiya Helper - dealing with multiple airline agents simultaneously to rebook a new flight. This storyboard will put the promised features and potential interface into a contextual story to showcase the usefulness.
Hiya Helper is here!
For Hiya Helper design process, we practiced rapid iteration, by iterating feedback from small scale usability testing among other colleagues, family member, and friends. This method allows us to fit into the tight timeline while implementing as many user-centered design as possible.
Full Design Documentation ->
Concept Report
Go ahead! Try Hiya Helper yourself!
The deliverable for the project is a design handoff.
As a team, we have given a full presentation to Hiya to deliver our concept and showcased our prototype. We handed off all the documentations and design assets, hoping to see that Hiya develop the real product one day.
Reflection
This is the first time that I experienced a product pivot during the middle the project, and it is a good practice for me to understand what is the “proper cue” to formally pivot away from the original pathway. Yes, it is painful to think about all the research efforts and deeply synthesized insights are not applicable anymore, but choosing the right project track can involve many sacrifices. There are no infinite time or budget to “force out a way” when project runs into dead-end.
I have led discussions and progress meeting during this project, and sometimes I have to be assertive and decision maker when everyone else is “I am fine either way”. Luckily, I have worked with a group of very flexible and accommodating individuals, but sometimes being confrontational is necessary to push the project forward, especially in terms of confirming research questions and deciding between design versions.
Team Appreciation
Again, this project will not be possible without my wonderful teammates. Even though this is my portfolio and it is mostly about “what I have done”, I still want to give credits to them for all the efforts and accommodations.
Through this experience, we have also grown together outside of the professional realm. A “better team connection” is even more valuable than a “kick ass” project.
Picture at the Hiya office
Before the final presentation
Combining the color of our outfit together = Hiya Purple 🟪
Casual team lunch
Casual team happy hour