EverBright | solar financing platform
Cutting contract rejections from 55% to 34%
A rebuild that encouraged sales reps to enter accurate numbers instead of inflated ones
My role
Design lead (sole designer)
Timeline
Q4 2024 to Q3 2025
Team
1 PM, engineering team, 1 design reviewer
Skills
Product design, systems thinking, stakeholder management
EverBright | solar financing platform
Cutting contract rejections from 55% to 34%
A rebuild that encouraged sales reps to enter accurate numbers instead of inflated ones
My role
Design lead (sole designer)
Timeline
Q4 2024 to Q3 2025
Team
1 PM, engineering team, 1 design reviewer
Skills
Product design, systems thinking, stakeholder management
55% → 34%
Utility-bill rejection rate
I designed the rebuild and the guardrails that helped bring the rejection rate down by 21%.
2 days
To ship a surprise regulation
When a new California requirement landed without warning, our team was able to swiftly implement a solution due to updated design system components in React.
4 loopholes
Closed in the rebuild
Guardrails prevented inflated entered savings by addressing: projected usage, editable time-of-use, uncapped usage, and unverified rates.
55% → 34%
Utility-bill rejection rate
I designed the rebuild and the guardrails that helped bring the rejection rate down by 21%.
2 days
To ship a surprise regulation
When a new California requirement landed without warning, our team was able to swiftly implement a solution due to updated design system components in React.
4 loopholes
Closed in the rebuild
Guardrails prevented inflated entered savings by addressing: projected usage, editable time-of-use, uncapped usage, and unverified rates.
The stakes
A 25-year promise, entered on aging code
Financed solar is a 25-year commitment. The Usage Phase is the part of EverBright's platform where a sales rep enters the homeowner’s utility data, including the provider, rate schedule, usage, and cost. Savings projections are built from these inputs, meaning that if they're inaccurate, the homeowner never gets the savings they thought they were signing up for.
In April 2025, EverBright implemented stricter utility bill verification to protect homeowners and reduce legal exposure. Overnight, rejected submissions jumped from 16% to 68%. The Usage Phase sat at the center of the problem, and it was running on aging, half-migrated code where every update was slow and risky. Our mission became: how can we continue to protect homeowners, without making EverBright harder to sell than other financiers?
Utility-bill rejection rate through 2025
The spike and the climb back down
Verification launched in April 2025 and rejections spiked overnight. Training, policy, and the Usage Phase rebuild brought them back down to 34%.
16%
Before verification
68%
Verification launches
55%
Training + policy
34%
Rebuild + guardrails
Rejections hit three personas at once
Sales reps
Each rejection meant redoing paperwork and another trip back to the homeowner. Reps have options; most sales orgs work with several financiers, so they're incentivized to go with whoever makes it easier.
Homeowners
A rejection meant a stalled sale and another ask for paperwork. For someone already nervous about a 25-year decision, each delay was a reason to back out.
Deal processing
Our internal deal processing team was spending their time chasing corrections instead of moving deals forward.
The stakes
A 25-year promise, entered on aging code
Financed solar is a 25-year commitment. The Usage Phase is the part of EverBright's platform where a sales rep enters the homeowner’s utility data, including the provider, rate schedule, usage, and cost. Savings projections are built from these inputs, meaning that if they're inaccurate, the homeowner never gets the savings they thought they were signing up for.
In April 2025, EverBright implemented stricter utility bill verification to protect homeowners and reduce legal exposure. Overnight, rejected submissions jumped from 16% to 68%. The Usage Phase sat at the center of the problem, and it was running on aging, half-migrated code where every update was slow and risky. Our mission became: how can we continue to protect homeowners, without making EverBright harder to sell than other financiers?
Utility-bill rejection rate through 2025
The spike and the climb back down
Verification launched in April 2025 and rejections spiked overnight. Training, policy, and the Usage Phase rebuild brought them back down to 34%.
16%
Before verification
68%
Verification launches
55%
Training + policy
34%
Rebuild + guardrails
Rejections hit three personas at once
Sales reps
Each rejection meant redoing paperwork and another trip back to the homeowner. Reps have options; most sales orgs work with several financiers, so they're incentivized to go with whoever makes it easier.
Homeowners
A rejection meant a stalled sale and another ask for paperwork. For someone already nervous about a 25-year decision, each delay was a reason to back out.
Deal processing
Our internal deal processing team was spending their time chasing corrections instead of moving deals forward.
the problem
An evolving problem inside the same scope
My initial brief was to migrate the Usage Phase to React, align it with our newly updated design system, and make small targeted UX improvements—not undertake a full redesign.
When reducing utility bill rejections became an additional (and primary) goal, the problem became: how do we implement the updates in a way that substantially moved utility bill rejection numbers without blowing past the scope everyone had agreed to? That balance shaped the design decisions that followed.
How might we
…turn a technical conversion into something that reduces rejections, continues to protect homeowners, and doesn’t expand into a full redesign?
the problem
An evolving problem inside the same scope
My initial brief was to migrate the Usage Phase to React, align it with our newly updated design system, and make small targeted UX improvements—not undertake a full redesign.
When reducing utility bill rejections became an additional (and primary) goal, the problem became: how do we implement the updates in a way that substantially moved utility bill rejection numbers without blowing past the scope everyone had agreed to? That balance shaped the design decisions that followed.
How might we
…turn a technical conversion into something that reduces rejections, continues to protect homeowners, and doesn’t expand into a full redesign?
end-to-end Process
Process
1
Revisited recently conducted research
Field research I’d conducted with sales reps months earlier surfaced an insight that informed this project: reps frequently adjusted usage data to present the savings they believed homeowners wanted to see. That made the Usage Phase more than a routine data-entry step—it was a point where reps’ decisions could compromise the accuracy of homeowners’ savings estimates.
Insight: Sales reps often manipulate usage data to tell the homeowner what they view as a more accurate story. Since savings is a projection, they believe that they need to project the utility data as well.
2
Audited every component, then filled the gaps
I reviewed each legacy component to determine whether we could replace it with an existing design system component. Where no suitable component existed, I partnered with the design system team to create one.
Before: legacy CSV upload
After: redesigned Document Upload
3
Designed guardrails
The product team found that selecting the wrong utility or rate schedule accounted for roughly half of utility bill rejections, while usage data errors accounted for most of the rest. Working with product, engineering, and risk, I redesigned these inputs to make accurate information easier to enter. Although seemingly small changes to a form, they greatly impacted whether a homeowner's promised savings actually occurred.
What we added
Made accurate inputs easier
Placed historical usage inputs above the chart
Added typeahead search to help reps find the correct rate schedule
Separated usage and cost into two charts that reps could toggle between
Added assistance program inputs, such as CARE and FERA
What we locked down
Made usage data harder to inflate
Capped annual usage to ~3x typical usage
Removed time-of-use profiles
Removed the projected-usage feature
Restricted California rate selections to verified defaults
GUARDRAIL: ANNUAL USAGE CAP + in-line error messaging
GUARDRAILS: COMBOBOXES + READ-ONLY STATES
4
Adapted to a new regulatory requirement in two days
An unexpected California regulatory change suddenly required 12 months of verified interval usage data for projects to proceed. Any missing data needed a documented explanation. Adding this requirement to the old system would have been slow and difficult. Thanks to the rebuild, we were able to ship a compliant flow in about two days, making the requirement clear and capturing a reason when data was missing.
The rebuilt California compliance flow
5
Led design reviews through a team re-org
A reorganization disrupted our review process, leaving a backlog of pull requests from several developers. I created a shared tracker for bugs and UI fixes, led reviews on a tight timeline, and worked with product and engineering to prioritize fixes before release. After launch, I facilitated a retro to identify what we could improve for future releases.
end-to-end Process
Process
1
Revisited recently conducted research
Field research I’d conducted with sales reps months earlier surfaced an insight that informed this project: reps frequently adjusted usage data to present the savings they believed homeowners wanted to see. That made the Usage Phase more than a routine data-entry step—it was a point where reps’ decisions could compromise the accuracy of homeowners’ savings estimates.
Insight: Sales reps often manipulate usage data to tell the homeowner what they view as a more accurate story. Since savings is a projection, they believe that they need to project the utility data as well.
2
Audited every component, then filled the gaps
I reviewed each legacy component to determine whether we could replace it with an existing design system component. Where no suitable component existed, I partnered with the design system team to create one.
Before: legacy CSV upload
After: redesigned Document Upload
3
Designed guardrails
The product team found that selecting the wrong utility or rate schedule accounted for roughly half of utility bill rejections, while usage data errors accounted for most of the rest. Working with product, engineering, and risk, I redesigned these inputs to make accurate information easier to enter. Although seemingly small changes to a form, they greatly impacted whether a homeowner's promised savings actually occurred.
What we added
Made accurate inputs easier
Placed historical usage inputs above the chart
Added typeahead search to help reps find the correct rate schedule
Separated usage and cost into two charts that reps could toggle between
Added assistance program inputs, such as CARE and FERA
What we locked down
Made usage data harder to inflate
Capped annual usage to ~3x typical usage
Removed time-of-use profiles
Removed the projected-usage feature
Restricted California rate selections to verified defaults
GUARDRAIL: ANNUAL USAGE CAP + in-line error messaging
GUARDRAILS: COMBOBOXES + READ-ONLY STATES
4
Adapted to a new regulatory requirement in two days
An unexpected California regulatory change suddenly required 12 months of verified interval usage data for projects to proceed. Any missing data needed a documented explanation. Adding this requirement to the old system would have been slow and difficult. Thanks to the rebuild, we were able to ship a compliant flow in about two days, making the requirement clear and capturing a reason when data was missing.
The rebuilt California compliance flow
5
Led design reviews through a team re-org
A reorganization disrupted our review process, leaving a backlog of pull requests from several developers. I created a shared tracker for bugs and UI fixes, led reviews on a tight timeline, and worked with product and engineering to prioritize fixes before release. After launch, I facilitated a retro to identify what we could improve for future releases.
Outcome
What changed
The Usage Phase rebuild improved data accuracy, reduced utility bill rejections, and made the system easier to adapt to changing regulations and market requirements. That meant fewer corrections for the deal processing team, less rework for sales reps, and more reliable savings estimates for homeowners.
Outcome
What changed
The Usage Phase rebuild improved data accuracy, reduced utility bill rejections, and made the system easier to adapt to changing regulations and market requirements. That meant fewer corrections for the deal processing team, less rework for sales reps, and more reliable savings estimates for homeowners.
learnings
Key takeaways
When a team changes, re-establish how you’ll work together
Before the reorganization, developers checked in regularly and I reviewed work throughout implementation. Afterward, reviews slipped to just before launch, leaving smaller UI fixes deprioritized under deadline pressure. Next time, I’d work with the new team to reestablish a regular design review process early on.
A stronger foundation makes room for the unexpected
The rebuild helped us respond to an unexpected regulatory change in about two days. Foundational work can be easy to undervalue, but it gives teams the flexibility to handle challenges they couldn’t have planned for.
learnings
Key takeaways
When a team changes, re-establish how you’ll work together
Before the reorganization, developers checked in regularly and I reviewed work throughout implementation. Afterward, reviews slipped to just before launch, leaving smaller UI fixes deprioritized under deadline pressure. Next time, I’d work with the new team to reestablish a regular design review process early on.
A stronger foundation makes room for the unexpected
The rebuild helped us respond to an unexpected regulatory change in about two days. Foundational work can be easy to undervalue, but it gives teams the flexibility to handle challenges they couldn’t have planned for.