Case Study: Capture

The Origin Story

In Dec 2022, contractors would email their project documents and photos. The the Program Team receiving the files would have to sort out what floor or room the work was done in and which photos went with which item. Since the utilities won’t issue rebates for a project they believe to be illegitimate, project documentation is key to the bottom line. They needed a way to get well-organized project completion packages.

This project, which came to be known as "DI Capture" was born.

Requirements

I met with the program team who receives the documentation to learn about their current process. When receiving documentation, they needed to know:

Location and equipment associated with the photos (building area and equipment type)
What specific photos are needed for each requirement
Proof the photos were taken on site and not manipulated

Pain Points

Their pain points included:

Long email chains with the contractors as the only way they managed projects
No consistent naming conventions for submitted files or photos
Time consuming to sort and organize materials upon receipt

How might we collect documents and photos systematically?

I met with the program team who receives the documentation to learn about their current process.

flow chart describing the rebate process and which stages require what.

My early understanding of the rebate application process and how an app could fit into it.

I thought about the mental model of a commercial building: Floors, rooms, and the energy-saving equipment that would be installed in each area. Using friction to slow contractors enough to confirm their location before taking a photo, I suspected we could reduce errors and confusion. And by assigning unique places for each required photo, the system could name them accurately and specifically.

Early ideas for MVP

Break down the photo list into areas matched with the floors and rooms of the project site, so contractors can drill down to an exact location
System assigns standardized file names automatically, based on area type and requirement
System geotags photos to verify location, so the program team can be more confident they’re not manipulated or reused from another job
Descriptive requirements listed for each photo, right in the app, so contractors have clear direction

I laid out some visuals in the company’s preferred whiteboard app, Whimsical, where the team could see and comment.

Four basic wireframes illustrating the flow of narrowing down the location and taking specific photos for each requirement

Research and Testing

In the interest of time, the team decided to forego user interviews and went straight to coding the MVP.

Test Environment

Screenshots of the early live app

Phone screen showing the landing page of the Capture app, with customer information blocked out

Landing page (customer info redacted)

Phone screen with compartments and add buttons for taking photos

Drilled in to the "measure" (equipment) level where each facet of the installation requests a photo

Contractor's camera view with instructions

Added Complexity

With the MVP out, we started to learn more about the complexity of the situation. During on-site testing, some new needs surfaced, related to communicating with the Program Team in the back office. Code violations, updates to the project status, and requests for change-orders, we learned, are all part of the day-to-day and needed to be accommodated.

A Quick Solution

Development responded quickly. Users are now able to add tags depending on the type of photo, in case they need to document a violation or request a change order.

Screen showing tag options, so that contractors can flag issues

Tags allow contractors to flag issues

A screen showing tags that label aisles like in a grocery store

Some projects, like grocery stores, are set up to categorize areas with tags for ease of sorting a large space

Contractor's camera view where they are about to document a Violation with a photo

More polished mockups

Figma components creating an example screen of what I thought the app would look like

I developed a more polished look for the tool in Figma. This was ultimately abandoned by the team in favor of using the color theming to match the other, more visually complex apps. Because of time constraints, no custom components were built.

Measuring Success

“DI Capture brought the 20% rejection rate down to less than 5%”

According to the pilot Program Team

Recommendations

Supporting translations, especially for photo descriptions, since many contractors are not native English speakers
Developing new, easily distinguishable components so that setting status and making change orders will look distinct from regular photo attachment slots

Thanks for Reading

I'm happy to go into more detail on this project on a call. Thanks for your time.

Project Documentation

My role: Concept, Wireframes