Remi
The Smarter way to organize your images
Grad School Project
Product Studio, Cornell Tech, Fall Semester 2021
Working period : Aug.2021 ~ Dec.2021
Collaborated with Cornell Tech grad students(MBA, CS, ECE)
for Product Studio's interdisciplinary team project
Partnered with
TEAM 64
- Nick Leonard / CS - Development
- Hoon Yoon / AAP - M.S.MDC - UI/UX/GUI Design
- May Olatoye / MBA - Product Planning
- Janet Liao / MBA - Product Planning
- Roberto Bruni / ECE - Development
How Might We enable mobile-first users in emerging markets to better manage their digital lives ?
Remi automatically aggregates event highlights between families and friends into rich, relivable memories. Users have the power to ‘ask Remi’ about event details and receive on demand responses and information.
Within developing countries, communicating daily through free messaging platforms provides the greatest source of digital activity for users. With an increasing number of event communication happening on these platforms, it creates a stack of digital assets such as images and comments that are often unorganized and not easily searchable.
Remi allows users to store, query and produce highlights of their group and event communications. By acting as a participant in group conversations, Remi parses, learns, and stores the groups interactions to provide easily searchable and shareable digital assets and responses.
Understanding user's digital life in emerging markets
In the research phase of Remi, through the initial user interviews regarding digital life in their emerging markets, we were able to map out what digital life looks like daily for the average emerging market consumer. We explored 3 markets across Brazil, Nigeria and India looking at aspects such as digital finance, to mobility, shopping, education and communication activities on platforms like Instagram, Facebook and WhatsApp. Through this phase, we found that within developing countries, communicating daily through the free messaging platform "WhatsApp" provides the greatest source of digital activity for users.
With an increasing number of event communication also happening on these platforms, it creates a stack of digital assets such as images and comments that are often unorganized, not easily searchable, and consume excessive amounts of storage and network bandwidth.
*The diagram graphics below is about comprehensive market analysis on daily digital life of emerging market consumer, and of which individual aspects are organized by several factors and segments.
And in the next stage, based on an initial market analysis diagram on emerging markets, we narrowed down our focus on a certain area in terms of opportunity sector: Nigeria for primary Market Targeting, WhatsApp for a basic service platform where we are going to investigate user's behavioral pattern and seek out the opportunity.
Value Creation Analysis
In materializing the project, ground on skeletal points for constructing fundamental concepts--which are the tools called "Value Creation Analysis"--we gradually developed our idea into more specified concepts and speculatively measured how we can actually create the product, interacting with the consumers in emerging market with securing retention rate of it.
In addition to basic Value Creation Analysis, we speculatively measured and calculated how we can possibly create profit from the market, based on comprehensive user data regarding user ratio, market share of WhatsApp(the platform we'll use as a bridge app to our product--Remi), approximate retention time on WhatsApp in emerging market--Nigeria.
* impression Rate: how many ads will user see per minute
Market Size Breakdown
Mobile device penetration rate in Nigeria: 48%
Population that have mobile device: ~7.1M
Social Media user: 15.8%
Percentage of user using WhatsApp: 93%
Remi Market Share (of daily active users): 5%
Daily Active User: ~50K
Avg time people spend on mobile: 38 min
Assuming showing 1 Interstitials ad per day (1440 min)
CPM Rates for Android Interstitials: $2-$4
TAM: 50k x (38/1440) x 2 x $1.5 x365 days= $1.44M
In designing the skeletal concepts for Remi, we also attempted to decompose our opportunity sectors into the matrix of Business Model Canvas, to clarify the elements that consist of the product structure, and to seek out the possible linkage between the elements.
In the last phase of designing concepts on Remi, we conducted an analytic tool--"The Optimal Derisking Sequence"--on our concept to clarify and derisk the any anticipable business threats(technical, business management-ish, society-related, cultural-related issues) that could seriously catch up with our future's business trajectory.
Creating personas
from research and actual users
Through the initial market research data targetting massive consumer group and actual user interviews--who are currently residing in developing countries, we were able to shape some coherent user data that comprehensively reflects the user tendency and traits on digital life in emerging market. Based on those extracted user data, we newly designed user persona and its user-scenario that represent the approximate user characteristics of emerging market, and by that, we intended to clarify how a person with a specific persona that inherently reflects our user analysis data could behave, live, and interact with her surroundings in digital life.
With creating user personas and its user-scenario, to concretize the core user experience of accessing Remi from the WhatsApp platform, we suggested a User Journey diagram that shows sequential phases.
Product Development
Experiment Planning based on Prototypes
With developing the product in the design side, in the user-testing phase we simultaneously devised the experimentation plans to measure the user response with actual users. One was chat experience simulation in WhatsApp's group chat room with acting like AI chat bot by team member, the another one was actual user testing with workable GUI prototypes of REMI product. For each experiment we plan to ask users a series of questions, on a one-to-ten scale, to try to measure things like speed, convenience, and excitement for the Remi.
The first experiment aims to simulate the chat experience with the Remi chat-like persona. The goal of this experiment is to capture the interaction and ease of use while trying to understand logistical information about an event with or without Remi as a resource to ask questions to.
The second experiment will simulate the rich, highlight-reel experience in the Remi App when compared to the vanilla WhatsApp scrolling experience.
Both experiments will focus on a sample of actual humans, not necessarily in emerging markets, to judge initial reactions first. These would be from a younger demographic and must be heavy users of WhatsApp already, meaning they participate in multiple group chats daily. For each experiment, one of the treatment groups will only have the usual WhatsApp experience to accomplish the task, while the second treatment group will have access to a Remi prototype connected to the WhatsApp experience.
Experiment Response
During 2 weeks, we met with 30 different users from emerging markets to demonstrate our prototype. Half of the users experienced WhatsApp only, and the other half experienced the Remi app.
Experiment Result
< Positive Outcomes >
-
Key features that strongly resonated with our Remi treatment group:
-
Improved search functionality
-
Auto-tagging ability
-
Limited impact to local device storage
-
Visually appealing highlight reel
-
-
90% of participants rated Remi’s summarization and aggregation features on top of WhatsApp as useful
-
“We love to store memories”
-
-
Freemium model is appealing
-
“This is not charging me for storage costs?! That’s great!”
-
Indicates that ad-based revenue model is likely the optimal monetization strategy
-
-
User can relate to event organization; i.e. this is an appealing wedge
< Shortcomings >
-
Users could see similar efficiency parallels with the photo highlight reel and similar reels their Android/Samsung devices generated, so value-add sometimes was minimal
-
Users prefer to see personalized content, not memes, gifs, or screenshots
Conclusion: Recommendations for Go-Forward Plan
Through the prototyping and its experiment sessions with participants, we have come to realize that our concept is positively responding to users' unmet needs of automatically storing and reorganizing group photos in WhatsApp group chat room. In particular, the technology harnessed in this project--personal AI-driven organizer that manages tremendous amounts of group photos based on extracted keywords from the message data--was what we thought as a compelling key point to those user group to some degree. To recap the feedbacks from test users, colleagues and faculties, most of them are commonly agreed to say that Remi could be developed into a tool for personal organization and mobile file management. The other feedbacks for Go-Forward plan for Remi are followed as below:
-
Test Remi’s potential within the small business and corporate space as an aggregation tool internally or a WhatsApp customer service content organization system
-
Further develop Remi’s business model and monetization strategy for emerging markets
-
Explicitly test for security and privacy concerns users’ may have w/ Remi
-
Ad based revenue model seems to be the best monetization strategy
-
Could we also user captcha/research based revenue model
-
We know that people would directly pay for this app
-
Run another experiment to better understand how big a barrier security and privacy are for this product
-
Run another experiment with an augmented product narrative focused on business users who utilize WhatsApp regularly to communicate with customers.
-
Chat based smart assistant hold value for certain use cases
Play the prototype version of Remi
TEAM 64
- Nick Leonard / CS - Development
- Hoon Yoon / AAP - M.S.MDC - UI/UX/GUI Design
- May Olatoye / MBA - Product Planning
- Janet Liao / MBA - Product Planning
- Roberto Bruni / ECE - Development
Remi
The Smarter way to organize your images
Grad School Project
Product Studio, Cornell Tech, Fall Semester 2021
Working period : Aug.2021 ~ Dec.2021
Collaborated with Cornell Tech grad students(MBA, CS, ECE)
for Product Studio's interdisciplinary team project
Partnered with
