Spotify for Artists Information Architecture and Navigation Redesign

Background

📚 Story Time

Long ago, in a not too distant past, Spotify had two separate tools: A tool for artists and a tool for labels. As the kingdom grew, the lines between artist needs and label needs blurred as new capabilities were released. With this growth there was a decision to merge the two tools into one — which was named: “Spotify for Artists (S4A)” — which ironically prioritized the needs of LABELS. Are you confused yet? Bear with me.

At the time, teams were so deeply focused on moving features into this new service, that there wasn’t a team (or even a consideration for that matter) on where features and toolsets should live. This responsibility fell on product teams who were already stretched thin getting their key features represented in this new service. Years passed.. Spotify grew into the behemoth it is today, and as new features and capabilities were released, this problem only became worse, the information architecture only became more messy. The pain our customers felt, only continued to exacerbate.

🤬 The Problem

Well if you look below at these gorgeous examples. Our information architecture (IA) and navigation are failing our customers. Our primary navigation is bursting at the seams — so much so that if you are on a small enough screen size, the content starts overlapping over other content. Our side navigation conflates items like artist views with global views, with user-level administrative actions. It’s a mess.

🙋🏻‍♂️ My Role

My role on this work was as the senior designer and the design owner for the Spotify for Artist platform team. I worked with one UX writer who collaborated with me on the content strategy and one senior UX researcher ran research with customers. I also worked with a senior PM.

Discovery 

👥 Understanding The Humans Who Use Our Service

To make things simple, we have 3 types of humans that use our service:

🔍 How We Uncovered Complexities

02 Workshops with internal teams

Bringing teams together to understand their roadmaps, and perform group card sort to better organize information with the experts.

04 Comprehensive Site flow diagram

Mapping out the entire flow of Spotify for artists as a tool to better help us understand our architecture.

🗂️ The Audit

We evaluated (nearly) every surface of the experience (both web and mobile) and graded how performant we are in when it comes to:

• Content hierarchy and clarity

• Effort required to use surface

• Relevance of surface to the workflow

• UX and copy inconsistencies

🗞️ Audit Findings

02 No easy way to manage multiple artists

Whether you are a giant label or an artist in your garage, we don't adapt the experience to your needs.

03 Disconnected omnichannel

No discernible pattern for which features are available on desktop web vs. mobile web vs. native apps. 

05 “You’re on your own”

New S4A users are left to poke around the site and discover things on their own. We don’t do much in the way of guidance.

It became clear that while we have many aspirations for our Spotify for Artist platform, we couldn’t achieve any of them without a clear Information architecture and navigation.

👯‍♀️ Understanding the needs of our many product teams

Takeaways:

• Many teams organized items into families that match key workflows our users are trying to achieve (promotion, engaging with fans, earning money, etc)

• Many sortings pointed to the need for a much more consolidated team management experience at the user level and not mixed in with artist views/actions

• All sortings that took place shared very little resemblance to  the current organizational structure

💚 Research with Customers

Getting feedback from as diverse a set of customers was pivotal to creating information architecture and navigation that resonated with all of our customers. There were several stages of research in which we were able to generate with customers and iterate off of their feedback.

I was fortunate enough to work closely with a sr. UX researcher throughout the many phases of research. She was the research lead on this portfolio of work. My role in research was helping craft our objectives and discussion guide; creating testing material such as mock ups, digital card sorts, and prototypes; finally, helping synthesize the information.

Stages

Stage 1: Generative Research

One initial step we took when performing qualitative research with customers was having them perform an open card sort of items in the current IA. This allowed us to quickly capture and better understand our customers mental models and key workflows. It was also a great jumping off point to talk with them about how they use our tools today and where they hope to see them evolve. Early on we also created tree tests where we would quickly and easily get a temperature check on how our team’s understanding and grouping of items matched our customers. This allowed us to move quicker and with greater certainty.

Stage 2: Survey sent to large sampling of customers

One other method we used to collect a wider range of mindsets from our customers was through surveying 1000 customers (.001% of user base). We wanted to understand what their top tasks in S4A were to ensure that our IA is as adaptive to those workflows as possible.

From the 1000 sampled, we found that the top tasks customers perform in S4A were:

• Release: Planning releases and boosting streams and growing awareness for music

• Analyze: Track and understand performance metrics

• Engage: Build fandom by engaging with their fans
  
  

Stage 3: Evaluative research (two tracks of valuation)

Current (top nav) v. side nav testing (1 round external + internal socialization)

Our research team ran tests with 8 Spotify for Artist customers. 4 of which represented a single artist and 4 of which represented medium to large labels. In these tests, the side navigation pattern was the clear winner. For labels they understood how it would create greater ease managing multiple artists and for single-artist teams they found the side navigation much simpler and straightforward to use.

Evaluative testing on new IA structure (2 rounds)

Across the two rounds of testing, the research team ran 22 sessions in total: 9 artist teams, 7 labels, and 6 managers. Our first round we aimed to validate the accuracy of our groupings and naming as it relates to their workflows. The second round was driving heavily by UX writing as it aimed to ensure that the specific naming had high resonance across all our many customer groups.

Exploration

🗒️ Concepts

The navigation went through many iterations. This process deconstructed and reconstructed certain elements and questioned and stress tested the reasoning behind decisions. What happens if we took away the top nav? — What are the essential workflows? — How can we decouple user level actions and artist level actions?

The fidelity started at sketches we used to quickly understand the world of directions we could go. When interacting with customers, we purposefully kept the fidelity low so we could focus the conversation on their workflows and have them engage with designs that purposefully felt a bit unfinished.

🔄 Switching from a top nav to a side nav

Many of our early iterations still featured a top nav as the main navigation. Through our collaboration with teams across our org and in understanding our customer’s workflows, it became very clear that relying on a top nav simply wouldn’t scale. In the coming two years, Spotify for Artists will grow meteorically, introducing new suites and families of tools. Even as we culled down and consolidated the current navigation, it was clear that the in its current state, the top nav had a clear expiration point where it would break. This also opened us up to use that space for more global features such as search, help, and user settings.

✏️ Collaboration with UX Writing

One of my main collaborators in this project was a mid-level UX writer. A huge component of this work is centered around content strategy; thinking critically how to create greater resonance with our content and families of content. Through our learnings we developed key heuristics for each layer of content in the IA — To share a sample, below is our guidance for primary navagation items:

Primary nav items

• A top-level nav item needs to encompass a unique suite of features, not just one tool

• No unique feature names, must be a familiar term

• Do not reflect internal org structures

• Optimize for an item that will be relevant and actionable for artists of all tiers

• Must be frequently used on a weekly basis, in all seasons

• To comply with accessibility standards and iOS/Android space limits, must be under ~10 character: Keep it simple.

Validation

🌊 Testing the Waters (10-50-100% rollout)

On any product, significant changes to the IA and nav can cause large customer upheaval. To reduce this risk we did a controlled rollout to a small set of customers, then to half, then to all. With any kind of IA/Nav redesign you actually expect to see a negative trend as customers adapt to the new landscape. To our surprise, we didn’t see any significant drop in engagement or number of actions taken. Qualitatively we also heard that people were finding the effort to find items had improved significantly.

🚀Final IA and Navigation

✅ Final Content in Primary Nav

Dashboard

The central landing for customers that shows their high level summary and learning material . This space will soon evolve into a space that surfaces relevant actions and information across S4A.

Fan Hub

Tools that help customers engage with fans.

Next Steps

👨🏽‍🎤 Artist dashboard

Now that we have an IA and navigation that can scale for our tooling well into Spotify’s future, the central focus will be on how to we better push content to customers. I am currently focused on reimagining the home experience for artists. Currently we use this page to push a lot of marketing material — and in talking to our customers, there is almost unanimous agreement that the surface they are dropped in when they log in is…completely useless.

This is a huge opportunity for our team to reimagine what “home” is for our customers. How can we better surface time-sensitive actions, useful analytics and faster entry points into their key workflows?

🪢Multi-artist home

In our prototypes for this phase of work we have already began diving into the multi-artist experience. The next phase involves adding a ton of fidelity to the explorations that we have begun and testing them with large labels.

📏 Creating templates

With the updates to the navigation, we have begun componentizing our designs so that they are reusable for the other product teams that build into Spotify for artists. This will include components for the nav with updated IA as well as an updated grid system to create greater consistency across product teams.