Song Designer

Personal project - 2 Months
UX Design - A songwriting app dedicated to helping musicians to capture, iterate, and get feedback on their ideas from anywhere.
Open InVision prototype
View pitch deck
Goal
There are 28 million musicians in the USA, but no dedicated songwriting tool. This means songwriters processes are inefficient and use many tools that aren't optimized for them, making it difficult to capture and finish musical ideas. How can we help users through their songwriting process?
Team and role
This was a solo project that started as my grad school capstone project. I conducted the research and formed initial concepts in 2019. Later in 2020 I picked the project back up to finish the designs, prototype, and begin to think as a product manager as I considered development.
Takeaways
Even when you are a target user, research is still important to avoid bias and find new opportunities you wouldn't have thought of yourself.

One of the hardest parts of creating a startup can be all the logistics and non-product related tasks that are required.

Try out the prototype!

Not everything is clickable, but you can try to:
  • Record a new audio idea
  • Write down a new lyrical idea
  • Edit an existing song
  • Add a new track
  • Add a new section
  • Try the rhyme, synonym, and chord keyboards

Research

Hypothesis

While in grad school, I made many parallels between the new design processes I was learning and the processes familiar to me as a musician. However, I also thought that musicians could learn a lot from design thinking, particularly when it came to writing songs. From my years personal experience, working with experiences musicians, and teaching new ones, I knew that we all weren't writing as many songs as we would like. I formed a hypothesis on why that was:
Many songwriters lack a defined or effective process or dedicated tools for going about their work. Their songwriting efforts could benefit from structure, and tools borrowed from the “design process.”
With a BA in music and a MS in design, I thought I was well positioned to design a tool that could drastically improves musicians' experience while writing songs. But while my hypothesis was based on professional experience, I started with research to validate my assumptions and uncover new insights that I may have otherwise missed.

Survey

I started with a survey and got 29 responses from a variety of musicians and songwriters. This helped me understand some general details about the type of musicians that would be interesting in songwriting, and gave me topics to follow up on in my interviews.
Key insight!
  • Phone recordings are the most commonly used songwriting tool, even beating the musician's primary instrument!
  • Phone notes are a close third, tied with the analog equivalent, pencil and paper.
  • Digital Audio Workstations (DAW’s) were beat slightly by vocals, but were still more commonly used than piano and other secondary instruments.
It became clear that I would be able to best help musicians if I went to where they were already writing. It was likely that I would be designing an iPhone application.
More survey details
Expert Craftsmen
Over 75% have more than 10 years experience.
Split Professionally
Many make their living through music, but many are enthusiasts.
Lyrics
74% write with lyrics at least half the time.
Popular Music
There is a focus on Pop, Rock, and Folk, which are quite similar stylistically.
Prolific Songwriters
Over 80% have written an album’s worth of songs, with nearly half having written over 30.
Constant Producers
Over 50% are writing a song nearly every month.

Musician interviews

With some of my basic assumptions validated, I needed to dig deeper into actual approaches that various musicians take when writing songs. Although I didn't get a diverse age group, I tried to get diversity in terms of experience, instrument, and genre.
Professional musicians
Sam
Bassist with 20 years experience
Professor, plays jazz and funk
Hannah
Bassist with 15 years experience
Student, plays jazz and folk
Non-professional musicians
Jacob
Guitarist with 10 years experience
Hobbyist, plays rock and electronic
Lincoln
Vocalist with 8 years experience
Soloist, plays pop and rock

Problem definition

Affinity diagram

To analyze my data, I wrote data points from my interviews on open response survey questions on sticky notes. As I arranged them, many categories and sub categories emerged.

Themes and insights

There was a lot of rich data that led to 10 insights across 3 broader themes. I could expand upon each bullet point at great length, but here is the high level.
Capturing
  • Musicians need a tool that can capture musical ideas anywhere, whether it is harmonic, rhythmic, or lyrical
  • Current musician tools are too "heavy weight" forcing users to choose tools that aren't optimized for them
  • Songwriters need to be able to listen back, loop, and combine/layer ideas to fully capture a musical thought
Developing and finishing
  • Loosely defining a familiar process could greatly improve productivity
  • Many writers stop after the initial idea, and could use support in moving forward, improvising and creating
  • Songs are systems, so musicians need to move quickly between all elements
  • Access to other songs and artists at the right time could help inspire and prevent getting stuck
Getting feedback
  • The are many questions of quality, direction, and variety on which musicians could use input
  • Songwriters frequently defined success in terms of other people, but never thought about them when writing
  • Digital Audio Workstations (DAW’s) were beat slightly by vocals, but were still more commonly used than piano and other secondary instruments

A songwriting story

Since most people aren't songwriters, in can help to get a sense of these insights in the context of a typical musicians journey of writing a song.

I mapped out a songwriting journey that is consistent with my personal experience as a musician, but was created from my research and is supported by user quotes.

How might we...

Help songwriters capture their musical ideas, work through a process to finish them, and iterate based on musician and fan feedback?

Ideation

Concept exploration

I began by generating ideas on the whiteboard so that I could go quick and explore many options. Because I had many insights, I would try to focus on only 1-2 at a time. To help, I would write the insight at the top of the whiteboard before starting. I would then star my favorite ideas, and start to combine them into more cohesive and workable concepts.
Song cards
I started by working on a list of song cards that would be familiar to users of voice memos and notes, since 80% plus of musicians interviewed are currently using those tools.
Song editing
But I wanted to take things a step further allow for editing across the many different parts of of a musical idea.

Concept test

Once I had the core workflow of capturing an idea (whether musical or lyrical) and saving it, I made a quick prototype to evaluate it. I did a quick task-based usability test with an experienced multi-instrumentalist musician. It had 3 tasks, and I followed up with questions about how he might use this.
The results validated that the initial concept was both:
  • Usable - He completed all tasks, commented on the easy of use, and enjoyed that it "felt familiar" to the tools he was already familiar with
  • Valuable - He saw a lot of benefit to organizing text and audio in one place, as well as some of the musician friendly features saying “if I just had rhymes right there, yeah that would be great!”
I also learned that he and others might actually use the text section for more than just lyrics. For example, he might want to include chords, versioning/iterations, or even just notes and related thoughts.

1.5 year pause...

At this point, the project actually took a pause for over a year. The capstone requirements for school were quite research and documentation heavy, so I never quite finished fleshing out the concept. With graduation, and moving to a new city for an exiting job, I left it in cold storage.

But the idea and passion never left me, and after months of the pandemic and isolation, I figured there was no better way to use my time than to pick it back up. My UI and prototyping skill had also greatly improved from my work experience, so I was ready to hit the ground running!

Hi-Fi iteration

After reviewing the research and basic workflows I had defined before, I moved towards higher fidelity exploration and more forward thinking features.
Song editor
The most complicated and essential part is the editor where musicians can move beyond what they have available in current tools to record tracks and sections.
Chord keyboards
Songwriters often just want to notate chords along with lyrics, but normal keyboards can make that tedious. It was fun to tackle how we might improve that.
Tips
It was clear from research that musicians often get stuck. I started exploring a future feature that could help push things forward in a variety of ways.

Design

Song list

The songs tab would look and function very similarly to the tools that a majority of musicians are familiar with, voice memos and notes. However, users would be able to playback audio with one click, and see different cards for songs that have audio, lyrics, or both.

The primary add song button is large, easy to reach with your thumb, and available at nearly all times. This makes song creation quick easy, and accessible.
Open InVision prototype

New song

Once clicking to create a song, the button to record audio is in exactly the same position and size, allowing musicians to learn over time that a quick double tap would start recording audio instantly. This is important because often times musicians come up with ideas while engaged in another activity (like driving... I know dangerous, but it is the reality).

One of the first core value propositions is that musicians can also capture lyrical ideas from the same place, simply by tapping the lyrics section just above.

Musician keyboards

One of the most valuable (if surprising) ways I found to deliver musician specific features that aided in songwriting was through the use of custom keyboards.
Rhyme and synonym
Often times when writing lyrics, musicians will consult dictionaries for rhymes when they are stuck on a line, or for synonyms when are feeling bland. These keyboards would auto select the word where your cursor is, and provide related words. Users could also tap different words above to get different recommendations, and could tap the recommendations to replace the original.
Chords
I know from personal experience, and validated through research, that notating chords with lyrics can be tedious. You have to switch between multiple keyboards to find letters, numbers, and symbols like sharps/flats/bar lines. This layout cleanly organizes exactly what a musician might need in a single view. There's also opportunity for ML chord recommendations in the future.
Key center
I leveraged my knowledge and passion for music theory to make an additional keyboard for those working in diatonic modalities. The top row of 3 being the blues progression, top two rows follow the circle of 5ths within the key center, and the smaller bottom row are common chromatic explorations for the key. All of this meaning that chords that work well together will be next to each other.

Full editor view

At any point, users can click a card on the song list and get taken into the full editor view. There is a tab for continuing to edit lyrics, and one for editing audio, which is empty in this case.

Notice that the add song button is still available, as I wanted to prevent users from having to navigate the app just to capture a new idea, since they so frequently come at random moments.

New track

Clicking to add a track brings up a menu where you can select some options for using a metronome and start recording. Once beginning the recording, you can also see the metronome visually which helps to prevent audio bleed if you aren't using headphones.

Selecting takes

After recording a take, users can review that take. They are also encouraged to try another version of that take with a rotating cycle of tips.

This process of ideation and iteration is one that is very familiar to musicians, but that is often looked over because none of the tools they use support it very well. Voice memos only records one idea, and the more professional digital audio workstations (DAWs) are more focused on perfecting than they are exploring.

Playback

After selecting a take, users see the full song view again. In this case, the user has actually recorded two tracks in the same section, meaning that those tracks will play together. This would commonly be used when a user had a chord progression they recorded and then the added a vocal melody on top, or for a bass riff and a drum groove. Users can mute and tracks, or click the edit button to select different takes or record new ones.

In addition, songwriters are often thinking in terms of sections, like verses and choruses. Clicking the new section button would allows them to record audio that is separate from any other audio they have so far, and would play right after the first section (or be moved later).

Future features

For my testing I wanted to include some hints (not fully clickable) of future features to get feedback. None of these would be included in an MVP production, but were helpful in crafting and updating my vision.
Filters
Of course one of the problems with voice memos is that all the quick one-off recorded ideas just pile up and get lost. One way of helping users get back to them would be to filter down to specific keys, tempos, or in the future even tags, that are related to their current work. Ideally some machine learning could help us identify keys and tempos without users having to manually enter them.
Tips and nudges
Another way of encouraging user to continue previous work is through a rotation of the many quotes from musicians saying so. This could even be a notification that nudges users back to the app, and links to their music for inspiration. In the future, I've imagined sections providing guidance and workflows on finishing songs or getting unstuck, potentially in a wizard-like flow.
Freemium
At some point I would want to start to monetize, but musicians generally only make $30-70K a year. Because of this, I believe the best approach would be to provide a free version for musicians to enjoy until they see enough value to upgrade. The free version might first be a limit of ~10 songs, and in the future limited access to more advanced features like the filters and tips.

Moving forward

Testing

Once I had a clickable prototype of these core workflows (with some future ideas mixed in), I wanted to do another round of testing. So far I have only tested with two people, but am in the process of planning a couple more! The current summary of results:
Went well
  • Immediately understands and values takes, tracks, and sections
  • Loved musician features like metronome, custom keyboards, and in app tips
  • Understood tips and feedback tabs just from landing page
Small improvements
  • The affordance for adding lyrics could be improved
  • The takes selection interaction in not intuitive or clear, and users might like to name or label a take with an icon
  • Unclear about input/output, could help to show status
  • Song summary bar could be improved
Requests
  • Collaboration (this was a big and common theme, something probably key to a medium term vision)
  • Folders to organize songs
  • Throwback Thursday or randomization for the “iterate” tip workflow
  • Import audio

Lean canvas

As my concepts became more complete and started to see validation in testing, I wanted to create a plan for how I might move forward. The lean canvas was a good way to quickly layout not only my core problem, solution, and value proposition, but how I intend to gain a user base and monetize. I have more details on all of the non-design aspects of the project, feel free to reach out and ask!

I'm actively looking for developers to work with on this project, so feel free to contact me if you are interested or have a referral!
Copyright Marshall Robbins - 2023