Image of Blossom prototype
Blossom is a smart device that helps plant owners take care of their plants. It makes use of both ambient light notifications and mobile push notifications to alert users to their plant's needs.
Research Methods: semi-structured interviews (remote and in-person), surveys, usability testing, prototyping, interface design
Contributions: interviews, surveys, sketching, storyboarding, physical prototyping, usability testing
Our initial research showed us that more young adults are becoming new plant owners. In the 2016 National Gardening report, it was found that 5 of the 6 million new American plant owners were between the ages of 18 and 34 (Bryant, 2017). In the 2018 report, it was estimated that 29% of all gardening households in the US fell into this age range (Globe Newswire, 2018). There has also been an uptick in the popularity of urban gardening in recent years. But while plant ownership has become more popular (Wenner, 2007), it's been estimated that approximately 32% of plants die within the first five months of purchase (Ellen & Komaromi, 2013). The fact that approximately one third of plants fail to thrive within the first few months of ownership was interesting to us and was an area we wanted to further investigate and see what possible solutions there might be to address it.​​​​​​​
We wanted to look into ways to help inexperienced plants owners feel better equipped to buy and maintain plants.
User Profile
From our early research, we established our initial user profile for our primary user group:
​​​​​​​Age: 18-34
Location: City
Income: Low to Medium
Technology Experience: Comfortable
Technology Available: Smartphone, laptop
Domain Knowledge: New to inexperienced
Attitudes and Values: Aesthetics, wellness
Semi-structured Interviews
Interviews were conducted both in-person and remotely, over phone or video call. We settled on semi-structured interviews to gain rich qualitative data that would be flexible as we were in the discovery stage of our research. We recruited and interviewed nine individuals who fit our user profile. We also spoke to three plant sellers to gain insight into what happens in stores as well as trends they've observed regarding the plant selection process.
The interview questions for our user group had slight variance depending on how an interview participant generally acquired their plant(s), whether they were purchased or gifted, and whether they currently had a plant or had plans to get one in the near future. The questions written to address our objectives were consistent among all participants in our user group. We tried to target participants who had purchased plants recently or in the past as well as those who planned to purchase some soon. Participants for our user interviews lived in different parts of the US and fit within our user profile.
Our main objectives in interviewing participants who fit within our primary user group were to find out:
     • why they choose to buy houseplants
     • what considerations do they have when buying a houseplant
     • where do they buy houseplants
     • how do they maintain their houseplants
Images of flowers.
The interviews for our plant sellers were conducted in two locations that specialized in selling plants in the Atlanta area. 
For our plant sellers, our main objectives were to find out:
     • what considerations customers tend to have when they buy a houseplant
     • what important factors customers should know before buying a houseplant
     • what important factors customers should know after buying a houseplant
     • the demographics of customers
     • the pain points of selling houseplants to customers
​​​​​​​Affinity Mapping​​​​​​​
We used qualitative analysis to identify key themes and issues from our interviews.
The main themes that came from our analysis were:
     • plant care is not obvious or easy to understand
     • plant owners sometimes don't know what plant they own
     • the information needed when buying a plant is not always accessible in one place
     • users have specific criteria that plant needs to fit into
     • there is emotional value in owning a plant
From the themes we identified, we began generating design ideas. These were then grouped in different  concepts that were graded on feasibility, relevance, and creativity. We then narrowed the concepts down to the three we felt were strongest and explored these options further.
Concept sketches for an AR application to help visualize plants in a space.
Our first concept was an augmented-reality app that would help users visualize plants in a location, find an ideal location for their plant, and give information on its ease of maintenance and care requirements.
Concept sketches for a kiosk to aid in plant selection and care.
Our second concept was an in-store kiosk that could suggest plants based on a user's criteria and give details about a plant, its care requirements, and optimal placement.
Concept sketches for a device that would track a plant's health.
Our third concept was a smart sensor that tracks and monitors a plant's health and would notify the user through ambient and mobile push notifications when the plant requires intervention.
These three concept focused on different aspects of the plant buying and maintenance experience. To decide on the direction we'd want to pursue further, we issued a survey to discover at which stage our users needed the most support and what pain points ranked as high priority. From the 39 responses to the survey, we found that respondents generally felt more confident buying a plant than they did in maintaining it. When purchasing plants, respondents wanted information on was whether a plant fit certain personal criteria (i.e. pet safe), how to best care for it, and where it should be placed in their home. For plants they already owned, the top three areas respondents needed help with were information on how to take care of the plant, how to diagnose issues the plant has, and how to identify when something is wrong with the plant.
Table mapping design concepts to features
Since the data from the survey seemed to favor something more care oriented, we decided to focus on our third concept, the soil sensor.
​​​​​​​This product has two components which we prototypedthe app and the sensor. For the app, we decided what user flows would be most important for the product and settled on the set-up, the viewing of a plant's health, and the addressing of a notification. The app screens were designed and a flow stitched together.
Set-up flow
Set-up flow
Plant health view flow
Plant health view flow
Notification flow
Notification flow
For the sensor, we focused on creating a mock-up, using Arduino to simulate the notification lights.
The form factor of the sensor was modeled after a Pycnosorus globus, or a billy ball. We decided to take inspiration from organic forms so that it would fit in with its plant and not be too distracting ​​​​​​​.
The project only allowed for a limited evaluation process, so we designed a discount usability test, having our subjects go through three task flows: to link the system to a plant, to view a plant's health status, and to address plant health notifications. We had them thinkaloud as they went through the three tasks, making sure they talked through their expectations and observations. As they tested our system, we collected notes on which aspects were most successful, aspects that were error prone or confusing, and the overall satisfaction level of our subjects at each stage. This method was chosen so we could observe real time behaviors and reactions. ​​​​
We had four subjects test our prototypes. They were recruited from our classmates and were current or former plant owners who considered themselves inexperienced in their plant ownership.
The subjects were given three tasks and as they went through them, we would prompt them with questions about what they would do next and what they expected to see from that action. Following the completion of each task, we followed up with additional questions specific to the task they had just completed. The follow-up questions were chosen to fit the task actions and track the user's thought process and whether a task was successful and what made it so or if not what led to the errors.
Task One - Set Up:
The subject was asked to link the app with the device and then associate the device with a particular plant.
Task Two - Plant Health:  
The subject was asked to view the status of a particular plant and communicate what information they understood from what they were shown. The user was also asked about the physical prototype, what they noticed about it, what they thought it meant, and what action they felt it called them to do.
Task Three - Notification:
The subject was presented with a push notification on a lock screen  that indicated their plant required attention and from there asked to find more information about it, execute the plant care action, and indicate that they completed the action within the app.
Task One - Set Up:
Overall, participants found the set up process easy to understand and complete, but there were some issues that came up:
     • confusion over the initial instruction screens 
     • unclear moment of device linking
     • lack of confidence in the auto-detection
     • unnoticed light notifications on soil sensor
Task Two - Plant Health:
Participants were able to find the appropriate pages to get information regarding their plant's health and felt that the information displayed was an appropriate amount and granularity,  but there was some confusion regarding the information that was displayed:
     • unclear meaning of bars in the graph
     • varying interpretations of colors in graphics
     • unnoticed status indicators
     • mismatch of personified copy and status indicators
Task Three - Notifications:
Our participants varied in their responses to the notifications, some preferring to go through the app and others acting as directed from the light notification. While each participant interpreted the app notification correctly, they each had different interpretations for the notifications from the sensor.
For future development, there are a number of things we'd like to change and further investigate. Most of the changes will address the confusion associated with the system, the set up process, and the notifications and notification elements. We also want to look more into the ambient notifications and how they can be improved.
Onboarding and Splash Screens
     • Summarize the features of the app and device and what the ambient notifications on the device mean.
     • Make copy more concise and clear, supplementing with visuals
     • Remove the button and make the screens swipeable.
Device Pairing
     • Outline steps for set up process
Onboarding and Splash Screens    
     • Include questionnaire about physical characteristics about plant 
     • Provide example image of suggested plant type
     • Rank plants in order of attention needed
     • Make notification indicator stand out more
Information Visualization
     • Label axes on graph
     • Provide key for color mappings and potentially change the colors used
     • Break down graph into the status categories provided
Mobile Push Notifications
     • Include action items that are currently be found within the plant’s status page
     • Allow users to address interventions from notification 
     • Bring user to plant's status page when notification selected
Ambient Notifications
     • Test granularity of information that can be conveyed through light notifications
     • Evaluate interpretations of colors
     • Have information about color mappings in the app
Bryant, T. (2017, March 21). Why Are Millennials Obsessed With Houseplants? Retrieved from Nylon:
Ellen, R., & Komaromi, R. (2013). Social exchange and vegetative propagation: An untold story of British
potted plants (Respond to this article at Anthropology
Today, 3-7.
Globe Newswire. (2018, April 18). Gardening Reaches an All Time High. Retrieved from Globe Newswire:
Wenner, M. (2007, April 17). Study Documents the Power of Indoor Plants. Retrieved from Live Science:

Back to Top