UX redesign for purchasing path and onboarding


Partial screenshot of the live Bullymake website redesign

“One suggestion for this company is to have an app that tracks were their buses are or to provide a list of estimated pick up times at their various locations. To wait for over 30 minutes at a stop, with children, only to end up standing and squeezed in is unacceptable.”
— DC Trails passenger, before this app was created


My client for this project is the CEO of Bullymake, a subscription box service for dogs.  They send toys and treats designed for “power chewers”, dogs who tend to chew through most other dog toys.

Bullymake is an e-commerce business which was founded in 2014.  Their main competitors are BarkBox and Amazon.  LootCrate used to offer a subscription box for pets but no longer does.  Bullymake stands out in their space because they only specialize in toys and treats for power chewers.  

Bullymake’s subscription service has experienced significant growth, especially within the past year.  However, they realized that there is always more opportunity to grow.  

Their CEO came to me because he wanted to improve his sites’ conversion rates.  He understood that his prospective customers’ largest objection was that Bullymake would always be the one to decide what would go in their boxes every month.  

Our work on his main site, bullymake.com, involved redesigning the product pages, purchasing path, and new subscriber onboarding so that either the subscriber or Bullymake could decide what kind of box a subscriber would receive.  My work as a UX consultant for Bullymake has involved creating provisional personas, creating mobile and desktop wireframes, and conducting UserTesting studies.

My task

  • Create provisional personas

  • Redesign the product pages and onboarding path, for both mobile and desktop

  • Conduct UserTesting studies

Project team

  • My role included UX design (in low fidelity) and usability testing.

  • Other members of Bullymake’s team included a graphic designer and developers. My work was handed off to them, and I reviewed the visual designs before they went live.




In longer engagements, I typically conduct user interviews or user surveys to find out ways that users vary in behavior and to create data-driven personas representing clusters of users.  For shorter projects like this one, I create provisional personas based on stakeholder interviews so that we have a shared understanding of who the users are.

For Bullymake, I created the following provisional personas:

  1. Ashley Jacobsen, an urban professional in her late 20s, who adopted a Boxer and takes him to dog parks.  She is a first-time dog owner.  She is a member of a wide range of other subscription services but has not tried one for her dog yet.

  2. Jeff Unger, a suburban professional in his late 30s.  He owns a bulldog, which he adopted when the previous owner moved, and previously owned another dog which was not a power chewer.  His dogs interact mainly with his children and play with toys in his yard or at a neighborhood park.  He prefers more traditional shopping and does not currently receive any subscription boxes.

  3. Ana Rodriguez, a suburban professional in her early 50s.  She bought one of her Labradors when it was a puppy and adopted her other one later.  She and her kids play with their dogs indoors or in their yard.  She has owned several other power chewers before, currently subscribes to BarkBox, and is trying to decide if she should switch her subscription over to Bullymake.

Designing for Jeff would have been a poor business choice because of his very low likelihood of joining a subscription service, but the site’s content could still offer something valuable to him.  So the only realistic design targets for Bullymake’s business were Ashley and Ana.  

I chose Ashley as the primary persona and the other two as secondaries.  Designing for Ashley would still make Ana happy, while designing for Ana could present some challenges to Ashley’s learning curve.

Qualification survey questions

Qualification survey questions

Screenshot 2018-03-19 15.25.15.png
Main survey questions

Main survey questions


Design iteration


From here, I created a list of 106 design ideas based on Bullymake’s needs, the personas’ needs, and inspiration from design galleries.  I deferred or did not select ideas that were too complex for implementation, required unplanned changes to Bullymake’s business, or were out of scope.  I prioritized the remaining ideas based on how well they fit the needs of the users, the business, and the available technology.

Next, I organized these ideas into pages for sketching so that my sketches and wireframes would accommodate all of them in a coherent flow.  Then I created these sketches and wireframes.  The wireframes included both mobile and desktop widths.

Screenshot 2018-03-19 15.54.08.png
Screenshot 2018-03-19 15.54.32.png
Screenshot 2018-03-19 15.54.44.png
Screenshot 2018-03-19 15.55.06.png
Screenshot 2018-03-19 15.55.25.png
Screenshot 2018-03-19 15.55.38.png
Screenshot 2018-03-19 15.55.51.png
Screenshot 2018-03-19 15.56.04.png
Screenshot 2018-03-19 15.56.21.png
Screenshot 2018-03-19 15.56.46.png
Screenshot 2018-03-19 15.56.59.png
Screenshot 2018-03-19 15.57.10.png
Screenshot 2018-03-19 15.57.24.png
Screenshot 2018-03-19 15.58.04.png
Screenshot 2018-03-19 15.58.19.png
Screenshot 2018-03-19 15.58.27.png
Screenshot 2018-03-19 15.58.52.png
Screenshot 2018-03-19 15.59.19.png
Screenshot 2018-03-19 15.59.33.png
Screenshot 2018-03-19 15.59.49.png
Screenshot 2018-03-19 16.00.29.png
Screenshot 2018-03-19 16.00.45.png

Usability testing

Screenshot 2018-03-19 15.52.38.png
Screenshot 2018-03-19 15.52.55.png

We conducted our usability studies through UserTesting.  The first one covered the mobile wireframes.

In the UserTesting screener, I made sure that we screened out testers who were “professional usability testers”, who do not subscribe to any subscription services, who do not own a power chewer dog.  Our testers owned a good representation of power chewer breeds.  Our test scenario involved looking for a new subscription service for toys for an 80-pound Boxer.   

The test itself involved comparing and contrasting two new proposed purchasing, signup, and onboarding flows.  UserTesting allowed us to give some participants variant A first and other participants variant B first, so we did this.  The main differences dealt with how we presented social proof and how many screens were part of the onboarding.  


Second design iteration


We decided to create two new concepts and do a desktop usability study.  Concept C was essentially concept A with B’s more category-based box builder.  Concept D started from concept B but pinned a Make My Box bar at the bottom of the product page so that people could sign up whenever they were ready.

Screenshot 2018-03-19 16.01.43.png
Screenshot 2018-03-19 16.02.21.png
Screenshot 2018-03-19 16.02.57.png

Live site

Thanks also to their team’s excellent visual design and development work, Bullymake launched their new site. By the time I took these screenshots several months later, their subscription base had grown from over 60,000 dogs to over 100,000 dogs!

Their team did a great job bringing the design ideas to life. Although there were some changes during visual design and development, the overall message, Pain/Dream/Fix landing page structure, and other design ideas from the wireframes remain in the live version.

Screenshot 2019-05-08 20.36.06.png
Screenshot 2019-05-08 20.37.06.png
Screenshot 2019-05-08 20.37.43.png
Screenshot 2019-05-08 20.37.30.png
Screenshot 2019-05-08 20.38.23.png
Screenshot 2019-05-08 20.28.57.png
Screenshot 2019-05-08 20.29.35.png
Screenshot 2019-05-08 20.33.52.png
Screenshot 2019-05-08 20.30.12.png
Screenshot 2019-05-08 20.29.47.png
Screenshot 2019-05-08 20.30.29.png
Screenshot 2019-05-08 20.31.47.png
Screenshot 2019-05-08 20.32.06.png



Recently launched - results forthcoming.


From the CEO

“David is extremely good at what he does. I highly recommend him for any UX work!”
— Colin Hatzmann, CEO, Bullymake

Let's work together.