From a link-sharing tool to a Today view

Part 1 · 2022–2024

What should the app do?

Starting out

Calendly pioneered the shareable scheduling link, and the company saw that link as the core of the product. The mobile app reflected that: it was mostly a place to copy and share your links.

The bigger problem was the business side: customers who adopted the app were more likely to churn than people who never touched it. The app had an identity crisis: if you asked three people what the app was for, you’d get three different answers. The realistic path was to just keep the lights on.

We had personas, but they weren’t based on real users, so I set them aside and went to see what people were actually doing.

Event types in 2021, each with Copy link and Share buttons. — Fig. 1The app when I joined in 2021: event types for copying and sharing links, and a basic list of scheduled events.

Scheduled events in 2021: a calendar and a list of upcoming meetings. — Fig. 1The app when I joined in 2021: event types for copying and sharing links, and a basic list of scheduled events.

Where were users finding value?

When I looked at the numbers, what stood out to me was scheduled events usage. It got a lot of use even though we’d given it very little love. Since every phone already comes with a dedicated calendar app, I was curious as to why users were spending so much time here.

Their answers were very insightful. They found their dedicated calendar apps were too noisy, full of other people’s PTO and blocks that often had little relevance to their day. Most Calendly users are professionals meeting external clients. Calendly was the filter on the meetings they actually got paid for.

Two behaviors kept coming up: booking the follow-up during the meeting, while rapport is highest, and checking what’s next and adjusting when a meeting runs long.

An interesting data point

Then I found a number that took me by surprise: about 1 in 5 events were booked by the host, not the invitee. Hosts were copying their own link, opening it in a browser, picking a time and typing in the other person’s details. That’s a workaround that went counter to Calendly’s ‘happy path’ of letting others book with you, and at that volume it meant we’d missed something people did with the product frequently.

The host’s own event types in the app. They tap “View scheduling page” on Chat with Tom. — Fig. 2About 1 in 5 events were booked by the host.

Their own booking page opens in a browser, and they pick January 9. — Fig. 2About 1 in 5 events were booked by the host.

Making the case for a new direction

In October 2022 I presented the finding to the whole product org at an offsite. Leadership pushed back hard. When I showed a video of a customer saying they booked 90% of their meetings this way, the room laughed. That night they put our senior data analysts on it, who came back saying that my numbers were conservative. That got me some credibility. The head of product became an ally and gave me the go-ahead to explore some new directions on mobile.

The reframe: we’d been building mobile like a smaller version of web. But people opened the app between meetings, managing what already lived on their calendar as often as they scheduled new events. The Calendly mobile app was for the real-time management of the meetings you get paid for.

I explored three directions. Leadership had suggested a low effort approach of making the scheduled events screen the home screen, but it buried event types and looked empty on light days. The focus view had the right priorities but was trying to do too many things at once. The split view, with relevant meetings up top and event types below, seemed to be striking the right balance.

Exploration: scheduled events as the home screen. — Fig. 3Three different directions I explored.

Exploration: a focus view with stacked sections. — Fig. 3Three different directions I explored.

Having a direction is one thing. Getting permission to build it is another. In early 2023 I tested it with about 40 users and looked at what bets were working and what needed refinement. Then I built the case one person at a time, up to the CEO. He approved it, but I don’t think he was convinced yet. That difference shaped the next three years.

MeProduct Designer
PM + Head of ProductInitial collaborators

New CPOFunded it, with more engineers
CPOBrought along

New PMJoined late 2023
CEOApproved

Fig. 4Selling the direction twice.

Getting approved didn’t make the project safe. In mid-2023, my PM, the head of product and the CPO all left the company. The new CPO came in with no history with the project, no relationship with me, and an open question about whether mobile deserved any more investment. So I made the case again from the beginning. He funded it, with more engineers than we’d had before. A new PM joined in late 2023, and we spent 2024 building it.

What we shipped in 2024

The 2021 home screen: event types with links to copy and share. — Fig. 5The home screen in 2021 and 2024. Each piece tied back to the research.

The 2024 home screen: a snapshot of meetings first, event types below. — Fig. 5The home screen in 2021 and 2024. Each piece tied back to the research.

Meeting details for a discovery call with Sam Goodman, with a Book follow-up button. — Fig. 6Booking on someone’s behalf: from a meeting, book the follow-up, pick a time from your full availability, and confirm.

When do you want to meet? Recommended times for the selected day. — Fig. 6Booking on someone’s behalf: from a meeting, book the follow-up, pick a time from your full availability, and confirm.

Results

In October 2024 the new home shipped alongside Calendly’s first in-app purchases. The headline: app users went from more likely to churn to more likely to stay and subscribe above web.

of mobile subscribers retained at 4th renewal (vs 57% on web): 65.6%of mobile subscribers retained at 4th renewal (vs 57% on web)
Paywall conversion (vs 8.76% on web): 19.2%Paywall conversion (vs 8.76% on web)
increase in traffic to scheduled events: +25%increase in traffic to scheduled events
boost in PSAT scores: +6%boost in PSAT scores

The 2024 mobile releases were among the company’s headline results that year, for a product line that had been a candidate for disinvestment three years earlier. The app was featured as App Store App of the Day several times.

Who was signing up?

About 48% of mobile subscribers were net new users who signed up and subscribed on mobile. Another 48% signed up on web and paid on mobile, and most of those, 35% of all mobile subscribers, paid more than 30 days after signing up. That means about 83% of mobile subscribers were net new to the company.

48%35%13%4%

Signed up and subscribed on mobile
Signed up on web, paid on mobile more than 30 days later
Signed up on web, paid on mobile within 30 days
Signed up on mobile, would likely have paid on web anyway (estimated from previous years)

Fig. 7Where mobile subscribers came from.

Part 2 · 2024–2026

Shifting towards day of execution

Where things stood

The 2024 home proved the thesis. By early 2026 the app was in good shape: retention had improved and PSAT was above 80%. We didn’t have a PM on mobile, so I set the roadmap. Three things shaped it:

According to customer feedback and surveys, users wanted more ways to engage and view their contacts.
Our CEO wanted more emphasis on meeting recaps and denser screens.
Mobile had never had a design system. The rules mostly lived in my head, and components drifted in code.

So we rebuilt the app around your day, moved scheduling to its own tab, and refreshed every style. The brand team sets Calendly’s colors and type. My job was how they come together on a phone: the layout and hierarchy, the details, how it feels to touch, and the system that keeps it consistent.

Three versions to the Today view

I started with a simple aim: make the home screen more useful for the current day. Each revision improved on the previous, and by the end it had become something new.

V2: holding a meeting card opens a menu with Join meeting, Reschedule and Cancel. — Fig. 8V1 to V3.

What shipped

The core shift: the app went from a home screen to a Today view.

The 2024 home screen: a meetings snapshot above event types. — Fig. 9The home screen in 2024 and the Today view in 2026. When there’s nothing to act on, the screen goes back to meetings and recaps.

The 2026 Today view: Today’s focus above the daily schedule, with recap thumbnails and short facts about each contact. — Fig. 9The home screen in 2024 and the Today view in 2026. When there’s nothing to act on, the screen goes back to meetings and recaps.

Mobile design system and AI tooling

You’ve been seeing pieces of the mobile design system throughout these screens. With AI changing how we build, a mobile design system became a 2026 priority. It’s net new work and about half my scope.

The 2024 contact screen: a large centered photo and name, Book meeting and Share availability buttons, then details, next meeting and previous meetings. — Fig. 10Contact and meeting details, before and after the design system.

The 2026 contact screen: a compact header with Schedule, Email and Phone actions, and an Activity tab listing notes, meetings and emails by day. — Fig. 10Contact and meeting details, before and after the design system.

Choosing where to start. I scored candidate components on impact (how often it’s used), effort (how hard it is to design and build) and priority (how much design and engineering needed it). The long-term goal is more than 50, but they don’t carry equal weight, so we started with the basics a coding agent needs to run with.

Checking quality

I make every design decision. The tooling I built checks the work against my bar.

A component audit. I turned our 60-plus item review checklist into a skill that runs in Claude Code or Figma Make. It tests things like dark mode contrast, 200% text and long translated strings.

9:41

Title

The component

Title, icon and button border against the dark surface.
The title at twice its size. Does it still fit?
A long German title. Does it get cut off?

Runs in your browser, on the component above.

Fig. 11Our screen header, rebuilt in code. Press Start audit to simulate the check.

A pixel diff. Ensuring design to code parity is a different problem. Web uses Storybook to handle this, but Storybook does not have much mobile support, so I used to be comparing screenshots by hand. I worked with engineering on an audit that renders components from code, overlays them on my designs and flags anything over a 2% difference. This has captured bugs we might not otherwise have caught, such as hidden text stylings being applied from legacy code.

Design Implementation Difference

Fig. 12The primary button as designed and as built. The build lands on the design with a difference blend, so anything that matches goes black.

What I’d do differently

In hindsight, I should have designed the mobile design system earlier. A design systems team was intended to eventually take on mobile, but it was never prioritized. As a result, for too long consistency depended on me alone, and older parts of the app drifted. My learning is that when a standard only exists in my head, that’s my cue to build the system.