Same Unit Repair
How to design and launch a new line of business during a pandemic
▲200%
▲200%
▲200%
Increase in jobs run through the app YTD from launch
▼83%
▼83%
▼83%
Reduction in incorrect resolution codes requiring manual intervention.
Team
I was the lead designer and researcher on this project on a small team with 5 engineers and a Product manager. I collaborated with my pal Bodhi on research execution and ideation.
Problem
Asurion Field was a product that I worked on in 2019 and 2020 that was developed for us by Asurion In-Home Experts to deliver new devices to folks who’d broken theirs really badly. Like, I-can’t-fix-this-thing badly.
But, most of the company’s claims—62% in 2020—were actually just screen repairs.
The question was this: How might we enable our in-home network of experts to help customers in a new way—by repairing their phones.
Ubreakifix was already in this space.
We had acquired Ubreakifix (UBIF) in 2018, and some markets had already started running jobs with remote repair technicians. We tapped into their knowledge to figure out how to incorporate screen repair jobs into our app for our internal users.
One big issue they were having with the program though, was that techs were often accidentally selecting the wrong job outcomes for customers (honest mistakes), which resulted in a lot of manual interventions to correct. Prior to this project, those UBIF jobs were coming in with the wrong resolution code 12% of the time.
We would need to fix that to scale this thing.
Asurion Field was a product that I worked on in 2019 and 2020 that was developed for us by Asurion In-Home Experts to deliver new devices to folks who’d broken theirs really badly. Like, I-can’t-fix-this-thing badly.
But, most of the company’s claims—62% in 2020—were actually just screen repairs.
The question was this: How might we enable our in-home network of experts to help customers in a new way—by repairing their phones.
Ubreakifix was already in this space.
We had acquired Ubreakifix (UBIF) in 2018, and some markets had already started running jobs with remote repair technicians. We tapped into their knowledge to figure out how to incorporate screen repair jobs into our app for our internal users.
One big issue they were having with the program though, was that techs were often accidentally selecting the wrong job outcomes for customers (honest mistakes), which resulted in a lot of manual interventions to correct. Prior to this project, those UBIF jobs were coming in with the wrong resolution code 12% of the time.
We would need to fix that to scale this thing.
Asurion Field was a product that I worked on in 2019 and 2020 that was developed for us by Asurion In-Home Experts to deliver new devices to folks who’d broken theirs really badly. Like, I-can’t-fix-this-thing badly.
But, most of the company’s claims—62% in 2020—were actually just screen repairs.
The question was this: How might we enable our in-home network of experts to help customers in a new way—by repairing their phones.
Ubreakifix was already in this space.
We had acquired Ubreakifix (UBIF) in 2018, and some markets had already started running jobs with remote repair technicians. We tapped into their knowledge to figure out how to incorporate screen repair jobs into our app for our internal users.
One big issue they were having with the program though, was that techs were often accidentally selecting the wrong job outcomes for customers (honest mistakes), which resulted in a lot of manual interventions to correct. Prior to this project, those UBIF jobs were coming in with the wrong resolution code 12% of the time.
We would need to fix that to scale this thing.
Discovery
Usually, I would have jumped in a Tech's van, gotten eyes on the job, and talked to actual, human users who were on the job.
Instead, I bought a GoPro, found a willing tech, and had him complete a diary study for us so we could get eyes on it from a distance. We learned a lot about the challenges that go along with this kind of work, and it was a great way to understand some of the intricacies of the work.
We really leaned on that UBIF team during our discovery. This was the summer of 2020, so it’s not like we could just do research the way we normally did. We had so many zoom meetings, a bunch of on-the-go FaceTimes with users, and learned all kinds of new ways to learn what we needed to learn.
Lol yikes.
We pulled this screen from one of the gopro videos we got back from the tech we worked with. This was the terminal screen techs got when a job was cancelled or otherwise unable to be completed. It's easy to see why over 12% of jobs resulted in an incorrect resolution code, requiring manual intervention.
Usually, I would have jumped in a Tech's van, gotten eyes on the job, and talked to actual, human users who were on the job.
Instead, I bought a GoPro, found a willing tech, and had him complete a diary study for us so we could get eyes on it from a distance. We learned a lot about the challenges that go along with this kind of work, and it was a great way to understand some of the intricacies of the work.
We really leaned on that UBIF team during our discovery. This was the summer of 2020, so it’s not like we could just do research the way we normally did. We had so many zoom meetings, a bunch of on-the-go FaceTimes with users, and learned all kinds of new ways to learn what we needed to learn.
Lol yikes.
We pulled this screen from one of the gopro videos we got back from the tech we worked with. This was the terminal screen techs got when a job was cancelled or otherwise unable to be completed. It's easy to see why over 12% of jobs resulted in an incorrect resolution code, requiring manual intervention.
Usually, I would have jumped in a Tech's van, gotten eyes on the job, and talked to actual, human users who were on the job.
Instead, I bought a GoPro, found a willing tech, and had him complete a diary study for us so we could get eyes on it from a distance. We learned a lot about the challenges that go along with this kind of work, and it was a great way to understand some of the intricacies of the work.
We really leaned on that UBIF team during our discovery. This was the summer of 2020, so it’s not like we could just do research the way we normally did. We had so many zoom meetings, a bunch of on-the-go FaceTimes with users, and learned all kinds of new ways to learn what we needed to learn.
Lol yikes.
We pulled this screen from one of the gopro videos we got back from the tech we worked with. This was the terminal screen techs got when a job was cancelled or otherwise unable to be completed. It's easy to see why over 12% of jobs resulted in an incorrect resolution code, requiring manual intervention.
Solution
We initially imagined that the screen repair job would be pretty linear, but we found out rather quickly that it wasn’t. There were way more outcomes to a job than we realized—it wasn’t as simple as “successful-” or “unsuccessful-repair”.
We tried to use a simple, linear path for techs, but after beta testing it, realized they weren’t able to reliably report back on the correct resolution code, despite their best efforts. They just had a lot going on in the field—adding to that cognitive load by asking them to select from one of 8 resolutions at the end of a job wasn’t helpful.
We tested a VERY linear user flow to start with, and realized quickly that it wasn’t going to work for our users.
Back to the drawing board
Instead of the linear flow, we added a series of questions into the flow that would branch, and lead our users to the correct resolution code during the job, so that they wouldn’t have to manually report it.
The second pass—a branching flow that prompted users to reach the correct resolution codes.
When we tested the branching flow, techs were so stoked. Technicians were frequently scolded for selecting the wrong resolution codes, and this took that weight off of them. It definitely took a bit longer to build than the linear flow, but removing the self-reporting aspect was a huge relief for them.
We initially imagined that the screen repair job would be pretty linear, but we found out rather quickly that it wasn’t. There were way more outcomes to a job than we realized—it wasn’t as simple as “successful-” or “unsuccessful-repair”.
We tried to use a simple, linear path for techs, but after beta testing it, realized they weren’t able to reliably report back on the correct resolution code, despite their best efforts. They just had a lot going on in the field—adding to that cognitive load by asking them to select from one of 8 resolutions at the end of a job wasn’t helpful.
We tested a VERY linear user flow to start with, and realized quickly that it wasn’t going to work for our users.
Back to the drawing board
Instead of the linear flow, we added a series of questions into the flow that would branch, and lead our users to the correct resolution code during the job, so that they wouldn’t have to manually report it.
The second pass—a branching flow that prompted users to reach the correct resolution codes.
When we tested the branching flow, techs were so stoked. Technicians were frequently scolded for selecting the wrong resolution codes, and this took that weight off of them. It definitely took a bit longer to build than the linear flow, but removing the self-reporting aspect was a huge relief for them.
We initially imagined that the screen repair job would be pretty linear, but we found out rather quickly that it wasn’t. There were way more outcomes to a job than we realized—it wasn’t as simple as “successful-” or “unsuccessful-repair”.
We tried to use a simple, linear path for techs, but after beta testing it, realized they weren’t able to reliably report back on the correct resolution code, despite their best efforts. They just had a lot going on in the field—adding to that cognitive load by asking them to select from one of 8 resolutions at the end of a job wasn’t helpful.
We tested a VERY linear user flow to start with, and realized quickly that it wasn’t going to work for our users.
Back to the drawing board
Instead of the linear flow, we added a series of questions into the flow that would branch, and lead our users to the correct resolution code during the job, so that they wouldn’t have to manually report it.
The second pass—a branching flow that prompted users to reach the correct resolution codes.
When we tested the branching flow, techs were so stoked. Technicians were frequently scolded for selecting the wrong resolution codes, and this took that weight off of them. It definitely took a bit longer to build than the linear flow, but removing the self-reporting aspect was a huge relief for them.
Impact
Scaling Screen Repair
The number of jobs we ran through the app increased dramatically after we launched same-unit repair in 2020. We went from 550,000 jobs in 2019 to more than 1,100,000 jobs in 2021.
Correct resolution codes
We also drove the number of incorrect resolution codes down from 12% to less than 2%. Overall, that's an 83% reduction in manual interventions on screen repair jobs.
Scaling Screen Repair
The number of jobs we ran through the app increased dramatically after we launched same-unit repair in 2020. We went from 550,000 jobs in 2019 to more than 1,100,000 jobs in 2021.
Correct resolution codes
We also drove the number of incorrect resolution codes down from 12% to less than 2%. Overall, that's an 83% reduction in manual interventions on screen repair jobs.
Scaling Screen Repair
The number of jobs we ran through the app increased dramatically after we launched same-unit repair in 2020. We went from 550,000 jobs in 2019 to more than 1,100,000 jobs in 2021.
Correct resolution codes
We also drove the number of incorrect resolution codes down from 12% to less than 2%. Overall, that's an 83% reduction in manual interventions on screen repair jobs.
Conclusion
This remains maybe my favorite project of my career. We had to get creative with our research, and ended up having a huge positive impact on the day-to-day life of our techs as well as the company’s bottom line.
And let me know if you ever want to run research for an entirely new product with a GoPro. I’m game to do it again any time.
This remains maybe my favorite project of my career. We had to get creative with our research, and ended up having a huge positive impact on the day-to-day life of our techs as well as the company’s bottom line.
And let me know if you ever want to run research for an entirely new product with a GoPro. I’m game to do it again any time.
This remains maybe my favorite project of my career. We had to get creative with our research, and ended up having a huge positive impact on the day-to-day life of our techs as well as the company’s bottom line.
And let me know if you ever want to run research for an entirely new product with a GoPro. I’m game to do it again any time.
