Phissy
After identifying a problem my grandmother faced, I designed a piece of assistive technology to address her needs. I had the opportunity to fuse my psycholinguistic and artistic training to bolster universal accessibility and appeal and to lead a diverse team to help mine consumer insight and inform our global expansion launch strategy.
Today, Phissy serves thousands of active users in over 40 countries, aged 18-85.
Phissy
After identifying a problem my grandmother faced, I designed a piece of assistive technology to address her needs. I had the opportunity to fuse my psycholinguistic and artistic training to bolster universal accessibility and appeal and to lead a diverse team to help mine consumer insight and inform our global expansion launch strategy.
Today, Phissy serves thousands of active users in over 40 countries, aged 18-85.
Phissy
After identifying a problem my grandmother faced, I designed a piece of assistive technology to address her needs. I had the opportunity to fuse my psycholinguistic and artistic training to bolster universal accessibility and appeal and to lead a diverse team to help mine consumer insight and inform our global expansion launch strategy.
Today, Phissy serves thousands of active users in over 40 countries, aged 18-85.
My Role
While I designed the product and wrote most code myself, I collaborated with other designers, software engineers, and researchers to build the framework of Phissy from scratch over the course of two years. Following its success, I’ve continued to oversee user research, design, and code-level optimization of our product’s usability and need fulfillment.
The Challenge
Grandma Phyllis “Phissy” Shaw is known for speaking her mind, particularly at restaurants, where dishes very rarely meet her elusive standards. I personally have witnessed dishes sent back again and again until the rest of us are on dessert. And next time we return to the same restaurant, the drama unfolds for an encore.
At the core of Grandma Phissy’s love-hate relationship with restaurants is not merely that she is what we might call a picky eater (which is to say a limitation of palate), but also a limitation of memory, and not one unique to her.
Maybe she has the boldness to vocalize her displeasure (and was not raised, as I was, on “you get what you get, and you don’t get upset”), but she also highlights a bigger problem: nobody can reliably remember everything they've liked and disliked at every restaurant.
Who among us have not asked,
The bottom line: we spend too much time ordering, eating, and judging restaurant food not to keep better and more efficient personal records of our experiences.
Discovery
Through weeks of consumer interviews and a competitive audit of nearly 100 apps in content-adjacent areas, it was clear where pre-Phissy consumer personae fell:
The Stenographer
Keep a shamelessly disorganized list (or lists) of all your dining history details in the Notes app or equivalent, then struggle to make any sense of it the next time.
The Hoarder
Write your notes on restaurant receipts and hold onto all of them in a junk drawer to reference… probably never?
The Influencer
Post photos and detailed public reviews of your meal. You might pause a meal to bust out a ring light just so you remember what you thought of the pesto six months later.
The Defeatist
Risk wasting time and money reordering something you didn't like the first time because it's easier than trying to keep track of everything you eat using the options above. (This was the most popular—by far.)
The food app space was saturated with having to rate restaurants by overall experience. This works for social media platforms like Yelp or delivery services like DoorDash, which rely on aggregate ratings to recommend new restaurants to potential diners. It doesn't work for food diaries and trackers.
Because it's not how we think.
Users were tired. They wanted easy, smart, full access to their own dining history to inform their dining futures. Sharing it should be a prerogative, not a prerequisite.
Phissy changed the game by making the dish the focus, enabling personal notekeeping with the optional flexibility to share with real people in real time—Phissy was the Google Docs of dining out.
Competitors that kept restaurant notes:
Competitive use cases:
Competitive usability:
Phissy’s MVP needed to disrupt the market by enabling users to:
Log what you ordered at a restaurant and what your friends ordered
Flexible customization of orders (additions, subtractions, reviews) as independent variables that can be parsed and indexed, not just a block of "notes"
Filter orders next time by rating, date, distance, cuisine, and more—or compound (e.g., "Alan" + "pizza" + 5 stars = Alan's all-time favorite pizza)
Share orders with friends and sync in real time without requiring you to post anything to the world
Development
Where restaurant-review apps’ focal unit was the restaurant, Phissy’s focal unit became the dish itself. This tree of nested folders (the Phissylist) proved most intuitive for navigating complex data:
I designed this flow to supplement memory loss gaps among older users. Episodic memory (which recalls experiences like dining out) is the first to decline in old age, unlike working memory, which temporarily holds information, such as a phone number while dialing.
A little brain science!
Episodic memory is built on a binding structure—for example, two variables must be linked to remember you once ordered eggplant parmigiana and had a positive experience with it: ([eggplant parmigiana]-[delicious]). More often than not, though, two-way binding is insufficient and three-way binding is necessary; say you’ve ordered the same dish at multiple restaurants with different reactions: ([Arturo’s]-[eggplant parmigiana]-[delicious]); ([Benvolio’s]-[eggplant parmigiana]-[awful]). Or one step further: ([Arturo’s]-[eggplant parmigiana]-[delicious]); ([Arturo’s]-[house cabernet]-[awful]); ([Benvolio’s]-[eggplant parmigiana]-[awful]); ([Benvolio’s]-[house cabernet]-[delicious]). Now six variables are in mixed association, each relying on another in the network to determine a given third. The dissolution of one binding tie sets the rest unraveling. The Phissylist is designed to mimic these structures, supplementing their binding pathways for individuals who had begun to lose their episodic memory.
How real people sort
Even when using a dish-centered app like Phissy, users still valued being able to sort restaurants by overall qualities. However, the content of these qualities differed.
Focus groups corroborated that when using restaurant-centered (rather than dish-centered) apps like Caviar, users are on the prowl for a delicious new gem in their area and care about:
distance
price point
cuisine type
When using Phissy, our research found, users are sifting through restaurants they may want to revisit based on their initial experience; they care about:
distance
when they visited the restaurant last
what their meal consisted of
how much they enjoyed it
This informed to the addition of dynamic search, sort, and filtering capabilities of past orders.
Achieving Product-Market Fit
When Phissy launched in spring 2021, consumers were emerging from COVID-conscious hibernation, both physically hungry and socially starved. I led a team from the United States, United Kingdom, and Italy—Phissy’s three top markets—to help monitor initial impressions of the application and assess how to build upon momentum and tap needs we initially had missed.
Our user base was skewing younger
Phissy's MVP was created for digital immigrants—older, less technologically literate adults for whom digital technology is primarily assistive. On the UI level, this had translated to:
minimal button quantity to reduce clutter and maximal button size to reduce error and frustration spikes
prioritizing legibility over originality in choice of typeface, adopting Apple’s native “San Francisco” typeface to facilitate cohesion with Apple’s native Notes, Mail, and Messages apps, with which users likely were familiar
opting for a greater than average font size contrast (i.e., the interval between title and subtitle font sizes), with the average font size three points larger than Apple’s and weight at least one point thicker
anything to do with “posting” or “feed” were stripped from Phissy 1.0, while “send a copy” and “sync in real time” social concepts were built in
But as downloads grew, we are surprised to see an exponential uptick in younger, digital native users who had tried the app and then given up on it. Raised with phones as limb extensions, this audience has distinct expectations for technology usability and functionality than our septuagenarian counterparts.
At Phissy, we had a new challenge—how to balance serving the intended function for older users with the expected function of younger users.
Key learnings → design principles
70% of our younger focus group participants wanted Phissy to become more playful, with a stronger presence of icons and images over text. However, iconicity needed to remain high—the interface could not become too abstract to make sense to older users.
The addition of in-app badges and rewards validated older users and enlivened the interface for younger users.
The color palette was remapped to dark greys with a pop of pink for verve—mature, but not outdated. (Standard dark-mode was a conscious decision; it makes the app less invasive to take out at even the classiest of restaurants. This is especially relevant for older crowds, who are sooner to frown at a glaring cell phone at the dinner table.)
We switched out the video tutorial for a built-in, interactive guided tour of the app that invites users to participate actively in learning at their own pace. (This does not require users who already have an episodic memory deficit to then remember every instruction given in a 12-minute video before entering the app for the first time!) At the same time, digital natives could bypass this guided tour instructions more swiftly if uninterested.
To the same effect, we paid careful attention to empty states. Since Phissy revolved around populating lists (each level of the Phissylist, the shortlist page, the collections page), I needed to craft an environment in which these pages, when unpopulated, would not intimidate to new users. To do this, I converted the friendly but unhelpful “There’s nothing here!” to concise directives, such as “When you add a restaurant to your Phissylist, you’ll see it here. Tap the plus button below to add your first restaurant!” This provides guidance to enable the older user to feel in control and the younger user to feel a much-anticipated sense of action.
To go social or not to go social
Too many apps have tried to be the next social media platform for food, a Instagram-Yelp hybrid. On paper, that’s compelling. The challenge with social is that if it doesn’t boom, it flops—no one wants to waste time in a virtual ghost town, and niches on the more established pillars (Instagram hashtags, Yelp Elite) steal market share from the less populous hybrid.
Since Phissy was designed to be a personal dining organizer for older users to comfortably providing the details of their orders without the fear of inadvertently publicizing it, this was a non-issue. However, younger users crave the social connectedness.
I circumvented the Instagram-Yelp pitfall by adopting a more Google Docs approach—users can foster a sense of community by syncing and sharing orders with friends who also have the app or by sharing and posting public restaurant reviews to their favorite social media platforms, ensuring no one has to post into a void.
The key mechanism behind this was Phissy AI. If users do want to post publicly, they can leverage A to turn their in-app dining notes instantly into an exportable review for use on any other platform. I trained six models to give users the flexibility to customize the voice of their review to match their own personal style and tone:
Results
4.8-star rating on Apple App Store among users in 40+ countries
60% user retention and growing
With special attention paid to cross-cultural and cross-generational language use, language analysis, intuitive and user-friendly design, positioning, and branding, we created a tool that solves an everyday problem with universal appeal, usability, and value.
What I learned
Studying the Apple Human Interface Guidelines equipped me to wireframe and build Phissy, as well as to conceptualize design and programmatic feasibility of future projects. I now am highly flexible in the Swift programming language and in NLP AI prompting.
As a business-oriented experience designer, where you end up isn’t always where you expect. Initially, I approached the task too narrowly and later had to reconcile two contrasting user needs: young users' craving honesty in technology and older, memory-impaired users' growing increasingly wary of it. More extensive preliminary research with a broader, more diverse sample could have minimized the need to pivot.
My Role
While I designed the product and wrote most code myself, I collaborated with other designers, software engineers, and researchers to build the framework of Phissy from scratch over the course of two years. Following its success, I’ve continued to oversee user research, design, and code-level optimization of our product’s usability and need fulfillment.
The Challenge
Grandma Phyllis “Phissy” Shaw is known for speaking her mind, particularly at restaurants, where dishes very rarely meet her elusive standards. I personally have witnessed dishes sent back again and again until the rest of us are on dessert. And next time we return to the same restaurant, the drama unfolds for an encore.
At the core of Grandma Phissy’s love-hate relationship with restaurants is not merely that she is what we might call a picky eater (which is to say a limitation of palate), but also a limitation of memory, and not one unique to her.
Maybe she has the boldness to vocalize her displeasure (and was not raised, as I was, on “you get what you get, and you don’t get upset”), but she also highlights a bigger problem: nobody can reliably remember everything they've liked and disliked at every restaurant.
Who among us have not asked,
The bottom line: we spend too much time ordering, eating, and judging restaurant food not to keep better and more efficient personal records of our experiences.
Discovery
Through weeks of consumer interviews and a competitive audit of nearly 100 apps in content-adjacent areas, it was clear where pre-Phissy consumer personae fell:
The Stenographer
Keep a shamelessly disorganized list (or lists) of all your dining history details in the Notes app or equivalent, then struggle to make any sense of it the next time.
The Hoarder
Write your notes on restaurant receipts and hold onto all of them in a junk drawer to reference… probably never?
The Influencer
Post photos and detailed public reviews of your meal. You might pause a meal to bust out a ring light just so you remember what you thought of the pesto six months later.
The Defeatist
Risk wasting time and money reordering something you didn't like the first time because it's easier than trying to keep track of everything you eat using the options above. (This was the most popular—by far.)
The food app space was saturated with having to rate restaurants by overall experience. This works for social media platforms like Yelp or delivery services like DoorDash, which rely on aggregate ratings to recommend new restaurants to potential diners. It doesn't work for food diaries and trackers.
Because it's not how we think.
Users were tired. They wanted easy, smart, full access to their own dining history to inform their dining futures. Sharing it should be a prerogative, not a prerequisite.
Phissy changed the game by making the dish the focus, enabling personal notekeeping with the optional flexibility to share with real people in real time—Phissy was the Google Docs of dining out.
Competitors that kept restaurant notes:
Competitive use cases:
Competitive usability:
Phissy’s MVP needed to disrupt the market by enabling users to:
Log what you ordered at a restaurant and what your friends ordered
Flexible customization of orders (additions, subtractions, reviews) as independent variables that can be parsed and indexed, not just a block of "notes"
Filter orders next time by rating, date, distance, cuisine, and more—or compound (e.g., "Alan" + "pizza" + 5 stars = Alan's all-time favorite pizza)
Share orders with friends and sync in real time without requiring you to post anything to the world
Development
Where restaurant-review apps’ focal unit was the restaurant, Phissy’s focal unit became the dish itself. This tree of nested folders (the Phissylist) proved most intuitive for navigating complex data:
I designed this flow to supplement memory loss gaps among older users. Episodic memory (which recalls experiences like dining out) is the first to decline in old age, unlike working memory, which temporarily holds information, such as a phone number while dialing.
A little brain science!
Episodic memory is built on a binding structure—for example, two variables must be linked to remember you once ordered eggplant parmigiana and had a positive experience with it: ([eggplant parmigiana]-[delicious]). More often than not, though, two-way binding is insufficient and three-way binding is necessary; say you’ve ordered the same dish at multiple restaurants with different reactions: ([Arturo’s]-[eggplant parmigiana]-[delicious]); ([Benvolio’s]-[eggplant parmigiana]-[awful]). Or one step further: ([Arturo’s]-[eggplant parmigiana]-[delicious]); ([Arturo’s]-[house cabernet]-[awful]); ([Benvolio’s]-[eggplant parmigiana]-[awful]); ([Benvolio’s]-[house cabernet]-[delicious]). Now six variables are in mixed association, each relying on another in the network to determine a given third. The dissolution of one binding tie sets the rest unraveling. The Phissylist is designed to mimic these structures, supplementing their binding pathways for individuals who had begun to lose their episodic memory.
How real people sort
Even when using a dish-centered app like Phissy, users still valued being able to sort restaurants by overall qualities. However, the content of these qualities differed.
Focus groups corroborated that when using restaurant-centered (rather than dish-centered) apps like Caviar, users are on the prowl for a delicious new gem in their area and care about:
distance
price point
cuisine type
When using Phissy, our research found, users are sifting through restaurants they may want to revisit based on their initial experience; they care about:
distance
when they visited the restaurant last
what their meal consisted of
how much they enjoyed it
This informed to the addition of dynamic search, sort, and filtering capabilities of past orders.
Achieving Product-Market Fit
When Phissy launched in spring 2021, consumers were emerging from COVID-conscious hibernation, both physically hungry and socially starved. I led a team from the United States, United Kingdom, and Italy—Phissy’s three top markets—to help monitor initial impressions of the application and assess how to build upon momentum and tap needs we initially had missed.
Our user base was skewing younger
Phissy's MVP was created for digital immigrants—older, less technologically literate adults for whom digital technology is primarily assistive. On the UI level, this had translated to:
minimal button quantity to reduce clutter and maximal button size to reduce error and frustration spikes
prioritizing legibility over originality in choice of typeface, adopting Apple’s native “San Francisco” typeface to facilitate cohesion with Apple’s native Notes, Mail, and Messages apps, with which users likely were familiar
opting for a greater than average font size contrast (i.e., the interval between title and subtitle font sizes), with the average font size three points larger than Apple’s and weight at least one point thicker
anything to do with “posting” or “feed” were stripped from Phissy 1.0, while “send a copy” and “sync in real time” social concepts were built in
But as downloads grew, we are surprised to see an exponential uptick in younger, digital native users who had tried the app and then given up on it. Raised with phones as limb extensions, this audience has distinct expectations for technology usability and functionality than our septuagenarian counterparts.
At Phissy, we had a new challenge—how to balance serving the intended function for older users with the expected function of younger users.
Key learnings → design principles
70% of our younger focus group participants wanted Phissy to become more playful, with a stronger presence of icons and images over text. However, iconicity needed to remain high—the interface could not become too abstract to make sense to older users.
The addition of in-app badges and rewards validated older users and enlivened the interface for younger users.
The color palette was remapped to dark greys with a pop of pink for verve—mature, but not outdated. (Standard dark-mode was a conscious decision; it makes the app less invasive to take out at even the classiest of restaurants. This is especially relevant for older crowds, who are sooner to frown at a glaring cell phone at the dinner table.)
We switched out the video tutorial for a built-in, interactive guided tour of the app that invites users to participate actively in learning at their own pace. (This does not require users who already have an episodic memory deficit to then remember every instruction given in a 12-minute video before entering the app for the first time!) At the same time, digital natives could bypass this guided tour instructions more swiftly if uninterested.
To the same effect, we paid careful attention to empty states. Since Phissy revolved around populating lists (each level of the Phissylist, the shortlist page, the collections page), I needed to craft an environment in which these pages, when unpopulated, would not intimidate to new users. To do this, I converted the friendly but unhelpful “There’s nothing here!” to concise directives, such as “When you add a restaurant to your Phissylist, you’ll see it here. Tap the plus button below to add your first restaurant!” This provides guidance to enable the older user to feel in control and the younger user to feel a much-anticipated sense of action.
To go social or not to go social
Too many apps have tried to be the next social media platform for food, a Instagram-Yelp hybrid. On paper, that’s compelling. The challenge with social is that if it doesn’t boom, it flops—no one wants to waste time in a virtual ghost town, and niches on the more established pillars (Instagram hashtags, Yelp Elite) steal market share from the less populous hybrid.
Since Phissy was designed to be a personal dining organizer for older users to comfortably providing the details of their orders without the fear of inadvertently publicizing it, this was a non-issue. However, younger users crave the social connectedness.
I circumvented the Instagram-Yelp pitfall by adopting a more Google Docs approach—users can foster a sense of community by syncing and sharing orders with friends who also have the app or by sharing and posting public restaurant reviews to their favorite social media platforms, ensuring no one has to post into a void.
The key mechanism behind this was Phissy AI. If users do want to post publicly, they can leverage A to turn their in-app dining notes instantly into an exportable review for use on any other platform. I trained six models to give users the flexibility to customize the voice of their review to match their own personal style and tone:
Results
4.8-star rating on Apple App Store among users in 40+ countries
60% user retention and growing
With special attention paid to cross-cultural and cross-generational language use, language analysis, intuitive and user-friendly design, positioning, and branding, we created a tool that solves an everyday problem with universal appeal, usability, and value.
What I learned
Studying the Apple Human Interface Guidelines equipped me to wireframe and build Phissy, as well as to conceptualize design and programmatic feasibility of future projects. I now am highly flexible in the Swift programming language and in NLP AI prompting.
As a business-oriented experience designer, where you end up isn’t always where you expect. Initially, I approached the task too narrowly and later had to reconcile two contrasting user needs: young users' craving honesty in technology and older, memory-impaired users' growing increasingly wary of it. More extensive preliminary research with a broader, more diverse sample could have minimized the need to pivot.
My Role
While I designed the product and wrote most code myself, I collaborated with other designers, software engineers, and researchers to build the framework of Phissy from scratch over the course of two years. Following its success, I’ve continued to oversee user research, design, and code-level optimization of our product’s usability and need fulfillment.
The Challenge
Grandma Phyllis “Phissy” Shaw is known for speaking her mind, particularly at restaurants, where dishes very rarely meet her elusive standards. I personally have witnessed dishes sent back again and again until the rest of us are on dessert. And next time we return to the same restaurant, the drama unfolds for an encore.
At the core of Grandma Phissy’s love-hate relationship with restaurants is not merely that she is what we might call a picky eater (which is to say a limitation of palate), but also a limitation of memory, and not one unique to her.
Maybe she has the boldness to vocalize her displeasure (and was not raised, as I was, on “you get what you get, and you don’t get upset”), but she also highlights a bigger problem: nobody can reliably remember everything they've liked and disliked at every restaurant.
Who among us have not asked,
The bottom line: we spend too much time ordering, eating, and judging restaurant food not to keep better and more efficient personal records of our experiences.
Discovery
Through weeks of consumer interviews and a competitive audit of nearly 100 apps in content-adjacent areas, it was clear where pre-Phissy consumer personae fell:
The Stenographer
Keep a shamelessly disorganized list (or lists) of all your dining history details in the Notes app or equivalent, then struggle to make any sense of it the next time.
The Hoarder
Write your notes on restaurant receipts and hold onto all of them in a junk drawer to reference… probably never?
The Influencer
Post photos and detailed public reviews of your meal. You might pause a meal to bust out a ring light just so you remember what you thought of the pesto six months later.
The Defeatist
Risk wasting time and money reordering something you didn't like the first time because it's easier than trying to keep track of everything you eat using the options above. (This was the most popular—by far.)
The food app space was saturated with having to rate restaurants by overall experience. This works for social media platforms like Yelp or delivery services like DoorDash, which rely on aggregate ratings to recommend new restaurants to potential diners. It doesn't work for food diaries and trackers.
Because it's not how we think.
Users were tired. They wanted easy, smart, full access to their own dining history to inform their dining futures. Sharing it should be a prerogative, not a prerequisite.
Phissy changed the game by making the dish the focus, enabling personal notekeeping with the optional flexibility to share with real people in real time—Phissy was the Google Docs of dining out.
Competitors that kept restaurant notes:
Competitive use cases:
Competitive usability:
Phissy’s MVP needed to disrupt the market by enabling users to:
Log what you ordered at a restaurant and what your friends ordered
Flexible customization of orders (additions, subtractions, reviews) as independent variables that can be parsed and indexed, not just a block of "notes"
Filter orders next time by rating, date, distance, cuisine, and more—or compound (e.g., "Alan" + "pizza" + 5 stars = Alan's all-time favorite pizza)
Share orders with friends and sync in real time without requiring you to post anything to the world
Development
Where restaurant-review apps’ focal unit was the restaurant, Phissy’s focal unit became the dish itself. This tree of nested folders (the Phissylist) proved most intuitive for navigating complex data:
I designed this flow to supplement memory loss gaps among older users. Episodic memory (which recalls experiences like dining out) is the first to decline in old age, unlike working memory, which temporarily holds information, such as a phone number while dialing.
A little brain science!
Episodic memory is built on a binding structure—for example, two variables must be linked to remember you once ordered eggplant parmigiana and had a positive experience with it: ([eggplant parmigiana]-[delicious]). More often than not, though, two-way binding is insufficient and three-way binding is necessary; say you’ve ordered the same dish at multiple restaurants with different reactions: ([Arturo’s]-[eggplant parmigiana]-[delicious]); ([Benvolio’s]-[eggplant parmigiana]-[awful]). Or one step further: ([Arturo’s]-[eggplant parmigiana]-[delicious]); ([Arturo’s]-[house cabernet]-[awful]); ([Benvolio’s]-[eggplant parmigiana]-[awful]); ([Benvolio’s]-[house cabernet]-[delicious]). Now six variables are in mixed association, each relying on another in the network to determine a given third. The dissolution of one binding tie sets the rest unraveling. The Phissylist is designed to mimic these structures, supplementing their binding pathways for individuals who had begun to lose their episodic memory.
How real people sort
Even when using a dish-centered app like Phissy, users still valued being able to sort restaurants by overall qualities. However, the content of these qualities differed.
Focus groups corroborated that when using restaurant-centered (rather than dish-centered) apps like Caviar, users are on the prowl for a delicious new gem in their area and care about:
distance
price point
cuisine type
When using Phissy, our research found, users are sifting through restaurants they may want to revisit based on their initial experience; they care about:
distance
when they visited the restaurant last
what their meal consisted of
how much they enjoyed it
This informed to the addition of dynamic search, sort, and filtering capabilities of past orders.
Achieving Product-Market Fit
When Phissy launched in spring 2021, consumers were emerging from COVID-conscious hibernation, both physically hungry and socially starved. I led a team from the United States, United Kingdom, and Italy—Phissy’s three top markets—to help monitor initial impressions of the application and assess how to build upon momentum and tap needs we initially had missed.
Our user base was skewing younger
Phissy's MVP was created for digital immigrants—older, less technologically literate adults for whom digital technology is primarily assistive. On the UI level, this had translated to:
minimal button quantity to reduce clutter and maximal button size to reduce error and frustration spikes
prioritizing legibility over originality in choice of typeface, adopting Apple’s native “San Francisco” typeface to facilitate cohesion with Apple’s native Notes, Mail, and Messages apps, with which users likely were familiar
opting for a greater than average font size contrast (i.e., the interval between title and subtitle font sizes), with the average font size three points larger than Apple’s and weight at least one point thicker
anything to do with “posting” or “feed” were stripped from Phissy 1.0, while “send a copy” and “sync in real time” social concepts were built in
But as downloads grew, we are surprised to see an exponential uptick in younger, digital native users who had tried the app and then given up on it. Raised with phones as limb extensions, this audience has distinct expectations for technology usability and functionality than our septuagenarian counterparts.
At Phissy, we had a new challenge—how to balance serving the intended function for older users with the expected function of younger users.
Key learnings → design principles
70% of our younger focus group participants wanted Phissy to become more playful, with a stronger presence of icons and images over text. However, iconicity needed to remain high—the interface could not become too abstract to make sense to older users.
The addition of in-app badges and rewards validated older users and enlivened the interface for younger users.
The color palette was remapped to dark greys with a pop of pink for verve—mature, but not outdated. (Standard dark-mode was a conscious decision; it makes the app less invasive to take out at even the classiest of restaurants. This is especially relevant for older crowds, who are sooner to frown at a glaring cell phone at the dinner table.)
We switched out the video tutorial for a built-in, interactive guided tour of the app that invites users to participate actively in learning at their own pace. (This does not require users who already have an episodic memory deficit to then remember every instruction given in a 12-minute video before entering the app for the first time!) At the same time, digital natives could bypass this guided tour instructions more swiftly if uninterested.
To the same effect, we paid careful attention to empty states. Since Phissy revolved around populating lists (each level of the Phissylist, the shortlist page, the collections page), I needed to craft an environment in which these pages, when unpopulated, would not intimidate to new users. To do this, I converted the friendly but unhelpful “There’s nothing here!” to concise directives, such as “When you add a restaurant to your Phissylist, you’ll see it here. Tap the plus button below to add your first restaurant!” This provides guidance to enable the older user to feel in control and the younger user to feel a much-anticipated sense of action.
To go social or not to go social
Too many apps have tried to be the next social media platform for food, a Instagram-Yelp hybrid. On paper, that’s compelling. The challenge with social is that if it doesn’t boom, it flops—no one wants to waste time in a virtual ghost town, and niches on the more established pillars (Instagram hashtags, Yelp Elite) steal market share from the less populous hybrid.
Since Phissy was designed to be a personal dining organizer for older users to comfortably providing the details of their orders without the fear of inadvertently publicizing it, this was a non-issue. However, younger users crave the social connectedness.
I circumvented the Instagram-Yelp pitfall by adopting a more Google Docs approach—users can foster a sense of community by syncing and sharing orders with friends who also have the app or by sharing and posting public restaurant reviews to their favorite social media platforms, ensuring no one has to post into a void.
The key mechanism behind this was Phissy AI. If users do want to post publicly, they can leverage A to turn their in-app dining notes instantly into an exportable review for use on any other platform. I trained six models to give users the flexibility to customize the voice of their review to match their own personal style and tone:
Results
4.8-star rating on Apple App Store among users in 40+ countries
60% user retention and growing
With special attention paid to cross-cultural and cross-generational language use, language analysis, intuitive and user-friendly design, positioning, and branding, we created a tool that solves an everyday problem with universal appeal, usability, and value.
What I learned
Studying the Apple Human Interface Guidelines equipped me to wireframe and build Phissy, as well as to conceptualize design and programmatic feasibility of future projects. I now am highly flexible in the Swift programming language and in NLP AI prompting.
As a business-oriented experience designer, where you end up isn’t always where you expect. Initially, I approached the task too narrowly and later had to reconcile two contrasting user needs: young users' craving honesty in technology and older, memory-impaired users' growing increasingly wary of it. More extensive preliminary research with a broader, more diverse sample could have minimized the need to pivot.