SmugMug · Collaboration & Account Management · Monetization | Zero-to-one | Complex workflows
I led design for Collaborator Logins from its first Beta to a paid product: a Beta to learn demand, an add-ons framework to sell it, and the seat logic to scale it.
ROLE
Lead Product Designer
TOOLS USED
Figma | Lovable | Miro | Google Suite| Quicksight | Zendesk
TEAM
Product | Engineering | Legal | Marketing | CX
TIMELINE
2025 - 2026
IMPACT
1.2K+ accounts adopted · $150K+ ARR run rate within 2 months of launch



● Release 1 · Beta (2025)
● Release 2 · Paid launch (2026)
● Release 3 · Enable / disable (2026)
My Role
Design lead from the first Beta to a paid add-on
I started by owning product and design, then led each release that followed, working closely with six partner teams.
_2x.png)
Background
SmugMug helps professional photographers store, show, and sell their photos. Many of them run busy studios with helpers: second shooters, editors, and office staff.
These helpers need to get into the photographer's account to do their jobs. But until this project, there was only one way to let them in.
The Problem
For photographers: one password, no control.
Every helper used the same shared "Assistant Password." Because of that, owners couldn't answer three simple questions:
-
Who can get into my account right now?
-
Can I remove just one person without kicking everyone else out?
-
Who changed what?
For SmugMug: nothing to sell to teams.
Growing studios wanted more, but a shared password gave SmugMug nothing to count, price or charge for.
At the same time, SmugMug was moving every login to a new system and starting to sell paid extras called add-ons. Collaborator Logins was going to be the very first one.
How might we?
How might I give each collaborator their own login, let owners stay in control, and turn team access into something SmugMug can sell—without making it feel complicated?
My Approach
Design for people, not seats
A "seat" is like a ticket on a bus. Software systems love thinking in seats, but photographers don't. They think: "I want to give this person access."
So I designed every screen around people first. I only brought up seats and billing when they directly changed what an owner could do.
My Design Process
Because our company ships on a two-week sprint cycle, I couldn't just drop this massive feature all at once. Plus, the backend engineering for this was deeply complex.
To get value into customers' hands early and give engineering the time they needed to build the core architecture safely, I broke the project down into four sequential parts. Each release solved a specific problem and laid the groundwork for the next.
Learn → Package → Scale → Fix
Release 1 · Beta (2025): learn what teams need
The question: How many collaborators does a studio really need? We didn't know yet, so we couldn't set a price.
What I did:
-
I framed the problem, scoped the first version, and got the team to agree on the plan.
-
We gave the Beta to about 2,000 Pro customers. That group was small enough for our support team to help everyone by hand.
-
I designed every owner and collaborator screen, from the invite email all the way to a collaborator's first sign-in.
How I showed the limit: Beta owners could only add a few collaborators. I tried three ways to show that limit:
-
A pop-up that blocks the screen. It’s clear, but it stops people right in the middle of a task.
-
A short status line on the page, like "1 of 1 seats used." It's always visible, but never in the way.
-
The status line plus a "Request more" button.
Why I chose it: The button opened a support form. Every time an owner clicked it, we instantly learned that someone wanted more seats. Each click gave us real demand data before we set a final price.
I also pushed to raise the Beta limit to 5 seats. If we kept it at just one seat, studios would just go right back to sharing a password.
-
Research highlights: Across the 2,000-customer beta group, ~450 owners clicked "Request more" to expand their limits, and ~120 support requests came into Zendesk during the first few weeks, giving us immediate qualitative feedback on how studios structured their teams.
SEAT LIMIT, RIGHT ON THE PAGE: Owners always see how many seats they've used, with a "Request more" button when they need more.

Release 1 · Beta: The limit sits right on the page. “Request more” opens a Zendesk form, so every click is a demand signal.

Release 1 · Inviting takes a name, an email and one consent check. On a trial, the path to more seats is “subscribe first”, not a dead end.
Release 2 · Paid launch (2026): turn it into a product
The question: Now that we knew teams wanted this, how should SmugMug sell it?
What I did: I led the new Add-ons framework, which is how SmugMug manages paid extras. Collaborator Logins went first. I made five key choices:
-
One page that figures out who you are: Instead of building four completely different checkout pages, I designed a single smart page. Whether you're brand new, already paying for Pro, testing a trial, or on an older plan, the page automatically figures out what you need to see.
-
Separating people from payments: I split this into two clean spaces. You manage your actual team on one page, and your credit card or billing details on another. That way, neither screen feels cluttered or confusing.
-
Keeping trials simple: If you're on a free trial, you get your starter seats for free. You can't buy extra ones until you officially subscribe, but the layout doesn't suddenly break or change shape when you finally upgrade.
-
Only asking when it matters: We didn't want annoying pop-up ads for more seats. Instead, the option to buy more only shows up the exact second you try to invite someone past your limit. It’s right there when you need it, and out of your way when you don't.
-
A shortcut for future features: Because we got these patterns right the first time, our team didn't have to reinvent the wheel for later add-ons like Source—they just reused these exact same screens.
ONE PAGE TO BUY SEATS: Every type of customer sees what's included and can add seats at $8 a month.
SIMPLE INVITES: A name, an email, and one consent checkbox are all it takes to invite a helper.
BILLING WHERE BILLING LIVES: Collaborator Logins sits right next to the main plan and other add-ons in Manage Subscription.

Release 2 · Paid launch:
One product page to buy seats: see what’s included, add seats at $8/month, then jump straight to Manage Collaborators.
Shown for a Pro customer; the same page adapts for new, non-Pro and trial customers.

Release 2 · Paid launch:
Add-ons live where billing already lives: Collaborators sits next to the plan and other add-ons.
Release 3 · Turn access on and off (2026): make it work for big teams
The question: What happens when a studio has a lot of helpers?
What sparked it: A studio with [20+] assistants showed us a new need. Owners wanted to pause someone's access—like a seasonal collaborator—without deleting them entirely. [Add how they fit inside the 20-seat limit.]
What I did: I tried five different layouts for the team list:
-
One flat list with an on/off switch on each row. (chosen one)
-
Separate "Active" and "Paused" sections. Rows jumped around unpredictably when you switched them.
-
Groups that open and close. This took way too many clicks to find a specific person.
-
A card for each person. This took up way too much screen space for a big team.
-
A dropdown menu on every row. Two clicks just for a simple on/off choice.
Why I chose it: Owners can see everyone's status at a glance and change it with a single click. Plus, it scales up cleanly as a team grows.
Small details that mattered: Over two rounds of feedback, I changed the button labels from action commands ("Enable access") to clear status states ("On"). I also added a "Saving…" message right after each click, error messages right on the exact row with a problem, and a clear cap of 20 active seats.

-
Release 3 · Enable / disable: pause access without removing anyone. A flat list with an inline toggle: status at a glance, one click to change it.
Release 4 · Seat logic ([shipped / in progress]): fix the count
The problem: Owners sometimes saw impossible math like "10 of 8 seats used." Invites were taking up seat counts, and turning someone back on never checked the limit, allowing numbers to go way over.
What I did: I worked through the logic with engineering in two rounds:
-
First, we stopped counting pending invites as active seats.
-
Then we found a second gap: when someone accepted an invite, the system automatically switched them on, once again skipping the seat check.
The rule we landed on: Sending an invite is free. Accepting an invite doesn't switch anyone on. Turning someone on is the only step that uses a seat.
Why it matters: SmugMug now bills correctly, and big studios can keep their whole team on a handy roster while only paying for the people actively switched on.
A SEAT COUNT YOU CAN TRUST: Only turning someone on uses a seat, so the count never goes over the limit.
How I worked with engineering
Shaped with engineering, not handed off
I didn't just hand off designs over the wall. Engineering and I sat down and decided these tough questions together:
-
What actually counts as a seat?
-
What happens to someone's access when an owner turns them off?
-
How does a free trial turn smoothly into a paid plan?
-
Where do extra purchased seats show up in billing?
-
How will named logins support activity history down the road?
I also mapped out every single path a person could take—including the messy edge cases: accepting an invite, declining it, clicking a broken link, clicking a link twice, and creating an account halfway through the flow.
Impact
Metrics:
1.2K+ accounts use Collaborator Logins
$150K+ ARR run rate within [2] months of launch
$8 a month for each extra seat—SmugMug's first-ever paid add-on
What changed
-
Owners got control. Each collaborator now has their own personal login, and owners can add, pause, or remove one person at a time.
-
SmugMug got a new product line. Team access finally turned into something scalable we could sell.
-
Future add-ons got a head start. Later add-ons reuse these exact same buying and managing patterns.
-
Growth has room. Named logins set the technical foundation for activity history and larger teams in the future.

WIP - Behind the scenes
What I learned
-
Design for how people actually think. People think about people and access, not "seats" and "billing." I made sure the system only appeared when it changed what someone could do.
-
Check every state of a feature you own, even the parts someone else designed before you joined the team. That's how I caught and fixed the "10 of 8 seats" glitch.
I took a shared password and turned it into a brand-new product line: handling the research, the pricing framework, and the design patterns that SmugMug's next add-ons now rely on.
