Master’s Capstone

This project was broken down into five main milestones detailed below.

Milestone 1 – The Idea


Pick-a-Pet allows shelters or pet fosters to safely provide pet trials for potential owners in public spaces. It also allows users that are unable to own a pet a chance to play with one and help the animal by providing play and exercise.

This would be a web-based application that could be accessed from any device or location. The application would show users what pets are near them and would provide information about the pet that would be obtained from a smart collar, microchip, or manual inputs. Information provided could include: pet type, breed, name, energy levels, age, photos, adoption agency, and location. This information will help users determine which type of pet would be best fit for their home or play time desired.

The goal of this application is to help shelters and foster homes adopt pets and provide play and exercise to the animals.

The Users
There are three potential user groups for this application.

    1. Pet lovers unable to adopt

  • A system that provides pets to play or exercise with for a short period of time in a public space without committing to adoption
  • These users may anyone that enjoys playing or exercising with an animal, but is unable to adopt one
  • This is a fairly large group, participants could be found at local parks or events
    2. Pet lovers looking to adopt

  • A system that provides pet trials in public spaces to ensure compatibility with a potential owner
  • These users could be anyone that is willing and able to adopt a pet
  • This is a fairly large group, participants could be found at pet adoption events or public parks
    3. Shelters or Fosters

  • A system that provides a safe way for potential owners to try out a pet and provides play and exercise to the pets
  • These users may be seeking potential adopters or may need help playing and exercising with the animals

Why Pick-a-Pet
I own three cats and would love to have a dog to take to the park and go for a run or play with, but my cats don’t get along well with dogs and we have a pet limit at the home we are renting. The public parks here in Houston are filled with dog-walkers and families playing with their pets and I thought it would be a great place to host pet adoptions or play dates for those that can’t adopt.

I don’t work for a company that does work related to this, but I could see it being integrated with smart collar applications and I do not have a team to work with. I chose this idea because it would help shelters/fosters and could provide some kind of companionship and play time to those that are unable to adopt.

Milestone 2 – Problem Space


Pick-a-Pet allows shelters or pet fosters to safely provide pet trials for potential owners in public spaces. It also allows users that are unable to own a pet a chance to play with one and help the animal by providing play and exercise.
Pick-a-Pet has two main stakeholder groups which includes rescue/shelter organizations that will generate information and pet enthusiasts/potential adopters that will view information. The latter group includes two sub groups that include users that may be looking to adopt an animal and users that love animals, but aren’t able to take on the full responsibility of pet ownership. This project will focus primarily on the users that will be viewing data with some focus on the groups providing data. Rescue and shelter organizations are quite busy and short staffed so connecting with the groups has been a challenge, but one user from this source has agreed to participate in the project and is considered a stakeholder. The user backgrounds are widely varied as the application appeals to many age groups, but there are a few commonalities among the participants interviewed. Some of the users currently owned pets, but didn’t own the type of animal they originally wanted due to the extra responsibility or due to a spouse bringing an animal into the relationship. Our participants are familiar with smartphones, tablets, and web-browsers and occupations varied from computer engineering, technical writing, to student. I connected with the rescue volunteer through Facebook and was able to interview the other group in person. I was able to meet with 5 end-user types ranging in ages from 24-56.

The rescues and shelters would use the application to post pet data, bring pets to public spaces, and collect data from pet-borrowers or adopters. The other user group would browse nearby pets, communicate with guardians, and post comments about their experiences with the animals.

The rescue worker may be working with multiple animals at once all at various stages throughout the rescue and adoption process. Rescue organization volunteers may not always have easy access to computers to create detailed posts about pets online and may have hard time tracking everything as they are often working in different locations outdoors while caring for or rescuing animals and indoors such as a veterinary office. Rescue workers are eager and in a hurry to find fosters or homes for animals as they are constantly getting new animals in and run out of space to house them. These users need to be able to quickly and easily create postings from anywhere and manage communications with prospects in one location.

The “renter” user group would utilize this app while out in public spaces such as parks to browse pets available for play nearby while the adopters may utilize the app while at home to see what pets are available nearby for adoption or foster. Both of these user groups may also browse pet data such as energy levels, active times, sleep time, weight, age, breed, etc…that could be pulled from smart collars or microchips. The renter users may have different reasons for borrowing a pet in a public space such as going on a date, getting some exercise, or satisfying the desire to play with an animal without taking on the responsibility of pet ownership.

Rescue/Shelter Tasks

  1. Receive rescue animal
  2. Provide care and medical services
  3. Begin search for foster or permanent homes
  4. Take photos of animals and collect information
  5. Create postings via social media, web sites, etc…
  6. Communicate with prospects
  7. Browse prospect profiles
  8. Make arrangements for meeting and possible drop-off

Renter/Adopter Tasks

  1. Browse available animals
  2. Filter through pets up for adoption, foster, play time, breed, location, etc…
  3. Communicate with guardians or rescue organizations
  4. Meet pets
  5. Possibly adopt or foster if interested
  6. Create profile

Rescue organizations use tools such as Petfinder, Facebook, pet stores, veterinary clinics, and Craigslist to create posts for available animals. These can be difficult to update and maintain as they are on multiple platforms and they may have many posts. Communication with prospects is also difficult from using multiple platforms. Rescue organizations need a single, mobile, tool to post, communicate, and track adoptions, fosters, and borrowers. If rescue organizations could utilize smart pet collars, entering data could be much more simple and prospects would be able to see relevant data to ensure they find the right pet for their lifestyle. However, smart collar technology is still fairly expensive, so the solution needs to helpful without the use of smart collars. Rescues also have a hard time matching the right pets with the right homes. This service could provide profiles with data collected upfront about prospects and pets that would provide better filtering capabilities to ensure that, e.g., high energy pets are matched with active homes and visa versa.
Pick-a-Pet will benefit prospects by providing recommended pet matches based on the profile data created and will allow borrowers to play with and interact with pets without taking on the full responsibility of pet ownership. Based on the user interviews, this could be beneficial for families with members who would like pets, but can’t own one, or for single adults looking for something fun to do while out on a date. These borrowers would also be providing much needed play time to the animals and can also get a feel for some of the responsibilities that could come with owning a pet. This service would also help parents teach their children about responsible pet ownership and what could be involved with owning one.


Rescue Organization




Pick-a-Pet success will be measured by providing the rescue and prospect participants with a high-fidelity click through prototype and follow-up survey questions using a Likert scale and collecting judgements.

Milestone 3 – The Design


Pick-a-Pet allows shelters or pet fosters to safely provide pet trials for potential owners in public spaces. It also allows users that are unable to own a pet a chance to play with one and help the animal by providing play and exercise. Users need to be able to browse pets nearby that are available for adoption or rent, they need to be able to view pet data and activity to see if the animal fits their lifestyle. They also need to be able to message the guardian of the pet to contact them for rental, adoption, or fostering. The application integrates with smart collars to ensure pet safety and tracking.


User Needs

  • Browse nearby pets
  • View pet properties
  • Contact guardian for pet playtime, adoption, or fostering

System Needs

  • Smart collar integration
  • GPS location and map of animals from smart collar
  • Filter results
  • Wi-fi and data messaging
  • Set virtual boundaries for pets via collar integration

Design Space

  • Microchip and Smart Collar Integration: I originally wanted to integrate microchip and smart collar data, but found that microchips don’t track GPS location or activity, microchips only hold a number that may be scanned to track pets within a system. The smart collar integration could still work, but the collars can run up to $450 which is not realistic for a shelter unless the collars are donated. To account for this, the application will need to work without the use of smart collars which may require the shelters to input some data manually. The current smart collars on the market don’t offer open APIs so much of the design will be conceptual for integration.
  • Filtering: Filtering capabilities may be difficult to keep simple while still providing complex options, but there are many existing applications that handle complex filtering well. There are also many filtering requirements to define such as adding functionality for users to match themselves with pets of similar activity levels.
  • Features vs. Functionality: I’ve had a hard time keeping features slim, as I work through the application I think of more features and capabilities to add or modify. I also keep in touch with the participants from the interviews and they provide more ideas as they come. I see a lot of different uses for apps similar to this one, but need to keep the design on track with the original intent to provide quality functionality without feature fatigue (Li, Wang, & Wu, 2014).

Julian lives in a house with roommates and can’t own a pet, but he loves dogs. He has a date and decides to take her to the park to walk around. Julian and his date ride the train and walk through the Japanese garden at Hermann park then take a break and watch all of the people out playing and running with their dogs. Julian decides it would be nice to walk around again and maybe play with a dog for a little bit with his date. He opens Pick-a-Pet to see what animals are available and close by then messages the guardian to come check the dog out. He follows the GPS to find the guardian with the dog and he and his date have a nice walk with the dog.

In my initial sketches I tried to figure out how to layout the navigation and how users should move through the product. I looked at apps such as Waze, Google Maps, Letgo, and Trulia to get some ideas.

sketch1 sketch2

I thought that for this design a layered approach would work as there aren’t many tasks a user would complete at one time. The sketch below shows an example of how the different areas of the application would be layered.

    • Purple: Map/Nearby Pets
    • reen: Menu
    • Teal: Messages
    • Red: Shelters
    • Orange: Profile


The map below shows an example of how these screens may branch off of each other

The layered approach provides users with only a few options at a time in order to avoid any confusion and to direct users to the primary tasks.

The main screen of the application shows a map with pins showing nearby pets. When a pin is selected, basic pet information will pop up from the bottom of the screen. Users may swipe this info up to view more or message the guardian or may select the map or swipe down to close it.

If the user swipes up to view more data, the pet info screen will expand vertically similar to a modal window and may be closed by selecting the “x.”

If the user messages the guardian, a new message modal window will open to allow for messaging. If the message window is closed the user will be taken back to the map screen.

The menu screen will open from the right side of the screen and lay over the map. Selections from the menu will open a modal screen similar to the Messages screen.

Future Technologies & Social Implications
Smart collars are fairly new and there are many types emerging that monitor and offer various features. Most of these collars integrate with a mobile application, but don’t offer open APIs yet to utilize the technology. I wanted to come up with a way to utilize the smart collar technology to benefit animal rescue and shelter organizations and I knew shelters were always looking for volunteers to come play with the animals to provide some exercise and attention so I thought this would be a great way to use the collar technology.

My husband and I love to go out to the park on a nice day and we always see people out running or playing with their dogs which makes me wish that I could have one to walk around. We can’t own a dog because we already have three cats and have a limit on animals from our landlord so I wanted to create an application that would also help people like us get some play time with a dog or even another cat.

There are some ethical issues that may arise from this service and application, with the biggest issue being safety of the animals and the guardians. Some of the smart collars like Pet Pace can measure stress levels by measuring heart rate, temperature, respiration, and position to ensure the pets are not being harmed. Other collars offer extensive tracking functionality and invisible barriers to ensure the pets are not taken out of the designated zones. The application will also need to verify user profiles via social media accounts or other information provided. Guardians may also need to collect additional data or get waivers signed when a pet is borrowed or rented.


  • Ming Li, Liya Wang & Mingxing Wu (2014) An integrated methodology for robustness analysis in feature fatigue problem, International Journal of Production Research, 52:20, 5985-5996, DOI: 10.1080/00207543.2014.895443
  • Smart Collars:

Milestone 4 – Prototype

Milestone 5 – Evaluation

I created a medium-fidelity click through prototype using Axure so that test participants can interact with the prototype on a mobile device as a realistic scenario. I am considering visual styles and themes for the application, but did not want to distract users by showing a finished looking product. Polished colors and styles can often mis-lead participants or create a bias based on visual preferences and perceptions. I considered using a paper prototype and possibly an app such as Prott or Invision, but I wanted to get a better feel for how the animations and interactions would work together so I chose to create the prototype with many of the intended interactions in place. Invision and Prott also create some issues with testing by providing hints to users after they make a mistake. Creating the prototype helped me adjust animations and navigation as I could click through it as I worked to preview animations. This prototype will also provide more details to define a rigorous usability study (Buley, 2013).

Because this prototype is still medium-fidelity with no styles applied it should be fairly easy to make adjustments based on usability test results. With limited time for the capstone course it is important to work lean and quickly to ensure an enjoyable, usable design is created. I don’t have a development team to work with, but my approach can still be lean as far as the iterative process goes and gathering user feedback throughout the project (Cyrillo, 2011). Once updates are made from usability test results, a higher-fidelity prototype may be created to test usability with visual styles which can validate the styles as well as the overall usability of the application. Ideally, I would test throughout each design phase, creating higher-fidelity mockups with each iteration (Wong, 2012). I did not formally test my sketches, but the prototype presented here underwent some design iteration from sketch to wireframe.

Evaluation Methods

Six participants were tested with various backgrounds and ages using a click-through prototype on an iPhone 6s Plus using Axure Share in a browser. The tests were conducted in a private area with the participant and myself as the moderator. Participants were given a test booklet containing nine pre-test demographic questions, three tasks, and a post-test questionnaire.

The confidence interval for completion rate was used to measure the usability of each task. Tasks that passed were given a score of 1, while failures were give a score of 2. Users were also asked to rate each task after completion on a scale from 1-7, 1 being “very difficult” and 7 being “very easy.” The confidence interval was used because this application is not being compared with another application nor is it being being tested against a benchmark. Task time was not a good measure for this study as task time may not validate navigation or completion. The Usability Metric for User Experience Scale (UMUX) (Bosley, 2013) was also used post-test to gather a short assessment that may be used to compare an updated version of the application in the future along with the scales gathered from each task and confidence interval. These evaluations allow for quantitative data that can be compared to future versions and provide stakeholders with numerical evidence of usability (Sauro/Lewis, 2012).

Introduction Script

Pick-A-Pet is a mobile application allowing animal rescue organizations to host adoption events and post pets available for adoption, foster, or rent. The application allows users to set preferences for desired pets. Pick-A-Pet also integrates with smart pet collars for safe tracking of rented animals and to collect activity data for potential adopters, fosters, or renters.

Thank you for your participation in the usability testing of Pick-A-Pet. This session will take approximately 15 minutes and will incorporate the following stages:

  1. Pre-test questions
  2. Demographic information
  3. Usability test tasks
  4. Post-test questions

General Information

  • Remember, the usability test is a test of the interface not of you. There is no right or wrong answer.
  • The software may be running on development machine, and/or the product may still be development. Please try to ignore technical and/or performance-related issues.
  • During the test your comments will be noted. Any information recorded will be used for research purposes only.
  • Your name will not be used in any reports. All information you provide will be reported anonymously.

Session Information

  • During the session you’ll be given three tasks. You will perform each task separately and only advance to the next task when the moderator asks you to do so.
  • Please read the entire task and ensure it is clear to you before performing it.

The Users

The participants for this study were potential pet adopters, fosters, or renters. The original intent was to test rescue organizations as well, but many constraints factored into testing with the organizations. The prototype was adjusted to accommodate the potential owners over organizations. In the future, this application would also need to be tested with organizations to ensure quick and easy pet updates and additions as well as creating events. The application will also need to be tested with various smart collar devices to ensure data is accurately being collected and analyzed.

Participant demographic and pre-test question data

Task 1 was difficult for every participant; wording, visuals, or layout may have factored into this task’s failure. Task 1 was written vaguely and was intended to move users through areas of the application without leading. Tasks 2 and 3 were to validate navigation and discoverability of messages and profile settings/pet preferences.


Task Ratings
Users were asked to rate tasks after completion from 1-7 with 1 being Very Difficult and 7 being Very Easy. Task 1 was only rated with 21 out of 42 possible points highlighting the difficulty or ambiguity of the task.

Task Confidence Interval
Tasks that were completed are scored with a 1, while tasks that were not completed without the assistance of the moderator are scored 0.
Both, the Confidence Interval and user ratings rate the application close to 80% whereas, it would ideally be closer to 90-95%. These scales show us which areas need improvement, it is clear that task 1 was the most difficult and unclear.

Post-Test UMUX Scale
The UMUX scale shows us that participants found the system to be somewhat frustrating. Some users noted that the prototyping application (Axure) was behaving oddly which may have factored into this result along with the difficulty of task 1.
Participants point out the difficulties with task 1 in the post-test, qualitative responses.


Overall, the application measured roughly 80% based on the UMUX, Confidence Interval, and user task ratings. The user task ratings scored 102 out of the possible 126 points based on easy of use which is 80%. The confidence interval actually gives us a higher percentage at 94% with the possible number of passes being 18 total, with 17 passed. However, most users passed task 1, but it was with much difficulty.


Based on the data, the highest priority issue would be to modify the events associated with task 1. The match notifications in task 1 were not obvious enough and users wanted to browse through adoption events to view pets. Some initial re-design ideas are to remove the match notification bubbles and add an extra layer to browse matched pets from the home screen rather than specific events or to create a high-fidelity screen to test again with users.

It was surprising that task 1 did so poorly, I expected the messaging or profile tasks to be more difficult. The messaging and profile portions may have done better due to them resembling familiar social media patterns.

Using Axure to test a mobile prototype may not be the best approach and I may not repeat it for future tests. The Axure prototype had a few bugs when viewed on a mobile device and never seemed to be very stable even on desktop. Too many risk factors are associated with an Axure prototype such as the browser and device. Although the Invision app is not ideal as it provides users hints after the first mistake, it may be better to use it or a similar application for mobile devices.

Works Cited

Bosley, J. J. (2013). Creating a Short Usability Metric for User Experience (UMUX) Scale. Interacting with Computers, 25(4), 317-319.

Buley, L. (2013). The user experience team of one: A research and design survival guide. Brooklyn, NY: Rosenfeld Media LLC.
Cyrillo, M. (2011). Lean UX: Rethink Development. Information Week, (1316), 40-44. Retrieved March 01, 2016, from

Sauro, J., & Lewis, J. R. (2012). Quantifying the user experience: Practical statistics for user research. Amsterdam: Elsevier/Morgan Kaufmann.

Wong, M., Khong, C., & Thwaites, H. (2012). Applied UX and UCD Design Process in Interface Design. Procedia – Social and Behavioral Sciences, 51, 703-708. Retrieved March 01, 2016, from