Built with AI
Rent increases (ECRI)
Storage operators raise rent on existing tenants regularly, sometimes for one unit and sometimes across several facilities at once. I designed the flow for scheduling those increases, for a single unit and in bulk.
Testing it. I was already running interviews for our leads research, so I added a usability test of the rent increase flow to the same sessions. The effective date sat at the top of the page, above the table, but people skipped right past it. They went straight to the table, filled in the new rent, and only noticed the date afterward, if at all. I checked this with other users too, and saw the same thing.
What changed. Putting the date first didn’t work, because the table is what people came to fill in. So I moved the date to the bottom, right before the Schedule button, in both the single-unit flow and the bulk flow. People fill in the table first, then set the date as the last step before they submit. Scheduling stays disabled until everything is filled in.
Approving increases. Some operators need someone to sign off before rent changes go out. By this point, PMs on the team were drafting designs in Claude Code, and a PM made the first version of an approval queue. I redesigned it. Increases can be queued for approval instead of scheduled right away, and the Queued tab lists each one with its proposed rent and its own effective date, so they can be edited, canceled, or approved one by one. Units don’t all have to change on the same day.
The first version of the rent increase flow shipped. The redesign with the date at the bottom was on the roadmap when I left.

The tested versions of the single-unit and bulk flows are on the password-protected page.View with password →
Built with AI
Leads: research and redesign
I took the initiative on this one. The leads page was designed before I joined, and as I used it, I saw room to improve it. The lead profile was one long form, and the notes, which are what people rely on most, were at the very bottom. When I had a lighter stretch between projects, I used it to research the problem.
Research. I wrote an interview guide and talked with four operators about how they work their leads. A few things came up again and again.
Findings and insights
- Notes and context matter mostPeople want to see who this person is and what happened last time before anything else.
- Required fields led to bad dataWhen the form asked for too much up front, people filled it in with placeholders just to save the lead. Name, a way to reach them, and a note were all they needed.
- Timing decides what to work on nextLeads go cold fast, and people wanted a clear next follow-up date that starts the clock.
- There was no quick way to tell which leads mattered mostManagers had to read every note to decide who to call first.
- Communication history was hard to findCalls, texts, and emails with a lead weren't easy to see from the profile.
- Managers wanted to see follow-throughThey wanted to know whether leads were being followed up on time.

Redesign. I built the redesign in Claude Code. I started with a quick, rough prototype to get the idea across, with communications on the first tab so the context people needed came first.
We couldn’t get feedback from clients directly, so I went to the customer success team. They’re the closest to our clients and hear what operators complain about every day. They told me the main job operators do on this page is filling out the lead form, so it needed to stay on the first tab. But the research was clear that people needed the important context right away too.
The answer was pinned notes. I added a row near the top of the profile for tasks and pinned notes, so the lead form stays on the first tab and the most important notes are still the first thing you see. Communications moved to the second tab.
I also designed a way to add an appointment date, for when a lead is coming in, right from the header at the top of the profile.
Built with AI
The pipeline view. A PM designed a Kanban board for leads, using the Claude Code prototyping process I’d brought to the team. I tightened up the design and added badges to the cards that show whether a lead has reserved a unit or is on the waitlist for one.
The redesign shipped.
The leads Kanban board is on the password-protected page.View with password →
Shaping how design worked at Monument
Setting up the process. After a team retreat, I wrote a proposal for how design should fit into the product process: design stays two sprints ahead of engineering, with a goal of a full quarter ahead. PM requirements turn into design tickets. Design gets involved as early as possible in discovery, with research methods matched to each stage, from interviews and journey maps to user testing. It also covered regular design reviews, a final review with the PM and tech lead before handoff, and a clear path for small one-off requests versus bigger initiatives. The team adopted the proposal, and it became how design and product worked together.

Built with AI
AI in how we design
Later, I moved our PMs and designers from Figma to prototyping in Claude Code, and led a Figma and Storybook audit of about 130 components so those prototypes used our own components. That work has its own case study: From Figma to Claude Code →
Promotions
Promotions are how storage operators fill units: “first month 50% off,” “third month free,” “$1 for your first month.” It sounds simple, but every operator runs promotions differently. Some start the discount at move-in and some months later. Some take a percentage off and some set a fixed rent. Some let promotions stack and some never do. I owned the design and the strategy, and the feature had to be flexible enough to handle all of it without becoming overwhelming to set up.
Discovery. The PM was on another project, so with the product director’s go-ahead, I ran the discovery myself. Every facility handled promotions differently. Some printed out the day’s promotions to keep at the desk. Others checked their email every morning. There was no easy way for staff to see which promotions were newest or how to pitch them to customers.
That led to one of my ideas for the feature: internal notes. Whoever sets up a promotion can add a note for staff, like “push this one first” or “customer must show a student ID,” and the note shows up wherever staff need it: in the promotion picker on the lead form, in the promotions table, and on eligible vacant units.
Testing V1. I talked with six operators, three facility or district managers and three directors of operations or revenue analysts, about how they create and apply promotions and walked them through my first design. A few things stood out.
Findings and insights
- Promotions and discounts are different things to operatorsA promotion is time-bound and expires. A discount, like a military discount, is a recurring credit for a specific customer.
- Manual setup leads to mistakesWhen someone forgets to apply a promotion, the fix is adding credits by hand every month.
- Our wording was confusingTerms like "duration-based" and "flat" didn't match how operators talk.
- Stacking was all over the placeSome operators allowed promotions to stack and others wouldn't, and most wanted control over which ones could combine.
- What workedThe side-by-side layout for setting up a promotion, the duration logic, and the internal notes.
From V1 to V2. I kept the side-by-side layout and reworked the rest around what I heard:
- Clearer choices up front. A promotion either uses a promo code or applies automatically, and it’s either an amount off rent or a fixed rent, like a $1 first month.
- A “starts on” setting, so a discount can begin at move-in or in a later month, which covers offers like “third month free.”
- Tiers for stacking. Stacking wasn’t a concept in the product yet, so I came up with tiers. Every promotion gets a tier, and tiers decide which promotions can combine. Operators who never stack can keep everything in a no-stacking tier.
- Display controls, so operators choose whether a promotion shows internally, in the online rental flow, or both.
- Occupancy ranges, so a promotion only applies when a unit type is within a set occupancy range.
- Notes where they’re needed. When staff pick a promotion for a lead, the options are grouped by tier, and each one shows its internal note right there.
Promotions shipped.

More from the platform
A few more screens from across the platform that I designed.
Unit groups and tiers. Operators can have hundreds of units at one facility. Unit groups organize them by size and type, and each group splits into three tiers, like Essential, Preferred, and Premium, each with its own price, amenities, and visibility in the online rental flow. The group page shows occupancy, pricing, and how rentals are moving between tiers.



Rates and rate plans. Rate plans adjust prices automatically based on occupancy, like raising rates 5% when a unit group is 50–70% full. The rates table lets operators review and change street and web rates for every tier, then apply all their changes at once.

Access control. Settings for a facility’s gates and keypads: opening access points remotely, grouping them into segments, setting access code length, and offering paid plans for extended access hours.

Adding amenities. Operators add the amenities a facility offers and pick an icon for each one.

Built with AI
Portfolio configuration. Operators set up their whole portfolio in one place: amenities, unit types, custom names for their tiers, and tenant and lead settings.
From my team
I hired and managed a product designer, and at different times managed two contract designers. I assigned the product designer’s projects and led his design reviews, and I set up the tenant portal’s components in Figma and the design system so he could make updates on his own. This project was his, with my direction and review.
Built with AI
Facility manager dashboard. One screen for a facility manager’s day, from priority tasks and past-due balances to today’s move-ins and new leads. I sketched the first wireframes a year earlier for client and sales conversations. He took it from there and built the final version with AI, and I reviewed it with him.
More screens, including portfolio configuration, storefront settings, and the facility manager dashboard, are on the password-protected page.View with password →