LANDR · 2020 · Senior Product Designer · ~2 weeks · Shipped · Measured
51% of new users left before hearing their first master. Making mastering faster would take months for under 5%, so I redesigned the wait instead: a progress bar built for perception, and a magazine to learn in the meantime. +79% songs mastered and $493K in incremental revenue.
Half of new users abandoned mastering during the Master Preview, the moment that turns a visitor into a paid customer: 71% of people who hear the preview go on to master a track.
The preview took about two minutes. Hotjar showed people leaving to do something else. In seven tests with new users, four found the wait frustrating and three thought it had frozen.
Engineering estimated months of work to make the preview less than 5% faster. Not worth it. So we redesigned the wait instead.
A five-stage breakdown replaced the spinner. Developers thought a simulated bar was impossible after earlier failed attempts. I found a way, and argued that perception mattered more than real-time accuracy. I estimated the average duration, sped up the bar for faster processes, and displayed a finalizing details message when they took longer.
Fresh production tips and other LANDR products, in a carousel that opens on its own. Leadership was reluctant until I drew a parallel with a dentist’s waiting room and demonstrated the potential revenue increase. The chance to cross-sell at a peak moment sealed it.
Many engaged users master several tracks in a row, and they hated the auto-trigger. We added a way to turn it off. The prototype testing we skipped under leadership pressure would have caught it earlier, but we still caught it.
LANDR is a one-stop platform for music production, and Mastering was its flagship. I worked across the whole Mastering journey, from the first upload to becoming a subscriber, and mapped it with the team: Sound DNA, the Waitzine, Revisions, pricing-page optimization, Synopsis. The Waitzine came first, at the point where the journey leaked most.
The Pro: a seasoned producer or mix engineer who works on music regularly and finds scattered production tools frustrating. Knows mastering well, and overestimates their expertise.
The Aspiring Musician: just starting out, learning as they go, looking for a platform that simplifies mastering and helps them progress.
Every decision had to work for both.
A new user picks a track, the Master Preview starts, the preview is ready, they create the master, they pay. 51% of users abandoned before the preview. With Mastering generating millions a year, that drop had cost millions in missed opportunities.
I went to the data, engineering and product teams, and asked for Hotjar. The engineer said the preview took about two minutes on average, and the data analyst later confirmed it. Hotjar showed many users leaving the platform before the preview, seemingly to do something else. The stakes were clear: 71% of people who hear the preview in the consideration phase go on to master at least one track.
I ran seven tests with new users from both personas, combining a cognitive walkthrough with performance testing. I watched task success, confidence, distractions during the wait, and how long the wait felt. Four of seven found it frustrating. Three thought the process had frozen. My hypothesis: some forgot to come back after something else pulled them away.
The obvious answer was a faster preview. I explored it with engineering; they estimated months of work to cut the time by less than 5%. We dropped it. The strategy became two things: progress feedback people could trust, and something worth looking at while they wait.
Leadership was reluctant to put content in front of people mid-task. Data and research got me a hearing. The CEO warmed to the idea when I compared it to a dentist's waiting room, where people pass the time with a magazine or their phone. What sealed it was the chance to cross-sell during a peak usage moment.
The spinner became a five-stage breakdown of the preview. It no longer looks frozen, and the bar reflects how long the process really takes. Developers resisted: earlier attempts had failed, and they thought it was impossible. I found a way to simulate the progress with high fidelity, and argued that user perception mattered more than real-time accuracy. They followed. People feel in control when they see progress, even simulated progress, and a smooth bar is easier to read than a spinning wheel.
Updating the progress panel alone fixed the feedback but didn't hold attention, so its improvements went into the final design. Full-screen feedback grabbed attention, but it would have bored people during the wait. And with the brand team and design-system leads, we chose a white top section over dark mode, so the progress would stand out.
With marketing, I balanced usability and content. At first we removed the buttons that led people away, to avoid more drop-off. Stale content would also weaken the Waitzine, so I ran workshops to push for monthly updates. It wasn't a marketing priority until they saw the impact a month later. Since then, CTAs have come back on every slide. I believe they hurt Mastering revenue, and I'd want Marketing, Finance and Product to compare overall revenue with Mastering revenue.
Songs mastered rose 79%, incremental revenue reached $493K, and cross-product adoption rose 6%. Showing other products mid-task was a risk we took with marketing and leadership: we wanted to avoid any drop-off, but supporting other products during LANDR's scaling period mattered. The result was positive.
We skipped prototype testing under leadership pressure. After launch, Pros cared less about the tips. Aspiring Musicians were the biggest winners, with many new users becoming engaged. Many users hated the auto-trigger because they master several tracks in a row, so we added a way to turn it off. What I took from it: building trust takes diplomacy, strategy and consistent execution, and in a leadership-driven room, data is what moves decisions.
Research, concept, interaction, UI, specs. With analytics, engineering, product, marketing, brand and leadership.
Prototype testing was skipped under leadership pressure. Marketing owned the content, later added counterintuitive CTAs to every slide. Noticed the change after I left.
I'd have fought harder to test before launch. The off switch fixed something testing would have caught.