Add a Feature: Spotify
Adding a feature to an existing, highly adopted app
I’m back, week 4 of Ironhack, and I only cried once. This week was our first individual challenge, where we were tasked to add a feature to a popular and highly adopted app in a 4-day sprint. I followed the Double Diamond design process — Discover, Define, Develop, and Deliver, — culminating in a high-fidelity prototype of my MVP, Minimum Viable Product. I actually really enjoyed this challenge, so let’s dive in.
Spotify was founded in 2006 in Stockholm, Sweden, by Daniel Ek and Martin Lorentzonhe and is now the world’s most popular audio streaming subscription service. They have a library of over 60 million tracks, 1.9 million podcasts, and 4 billion playlists. They currently have over 320 million users, including 144 million subscribers, and in over 92 markets.
Lean UX Canvas
I chose Spotify, mostly because it is possibly my favorite and most used app. and Music is life right? To get started I needed to conduct a Business Analysis to get a better scope of the company and its users. I entrusted my process with a Lean UX Canvas, which served as a framework for what I need to focus on and accomplish, by setting certain goals that would translate to user benefits. Normally, you would have access to the stakeholders however as that was not the case here, I made an educated assumption based on secondary research. With the Lean UX canvas, I was able to outline the Business Problem, possible Business Outcomes, defined the users I should focus on first, and User goals.
Competitive Feature Chart
Now that I have a framework to guide me, I needed to delve a little deeper into research. I started with the Competitive Feature Chart, which allowed me to discover who the direct (Apple Music, Amazon Music, Youtube Music) and indirect competition (Audible, Ted Talks, Live Shows) was. From there I started to list the features each had to uncover any patterns, mental models, and areas of opportunities there might be. It is crucial to include indirect competition because more often than not the company’s direct competition is in the same red ocean, or contested market space. In fact, the areas of opportunities I uncovered came from looking into the indirect competitors' features. But I do not want to ruin the surprise, so more specifics later!
Market Positioning Map
In order to analyze Spotify’s current position and competitors' position within their market, I turned to the Market Positioning Map, which allows you to visually determine blue oceans (uncontested market space). Based on the list of features created with the Comp. Feature chart, I looked to define the axis for the Map, to uncover possible Blue Oceans. I created a few however the map with the X-Axis, Niche & Popular, and Y-Axis, Multifaceted & Standard, provided the clearest blue ocean, to become more multifaceted thus becoming more popular.
Alright, so now I knew where Spotify sits within the market as well as who the competition is and where they sit and uncovered a blue ocean. Now I needed to talk to the targeted users of Spotify. I started by collecting Quantitative data via Surveys, which resulted in 34 responses. I based my questions on areas of opportunities I uncovered while using the Comp. Feature chart. I thought I could go a little longer without divulging too much so that you’d be blown away by the MVP, but that's not going to work. So the opportunity I uncovered was when I looked into the indirect competitor, Audible. Well, I guess I can leave it at that for now.
The main takeaways here were:
- 73% use Spotify on a daily basis
- 100% find Spotify easy to use
- 77% listen to Podcasts
- 67% use Spotify to listen to Podcasts
I now had a better understanding of the statistics associated with Spotify users. I wanted to better understand the perspectives and the behaviors by collecting Qualitative data in the form of interviews. I was able to conduct 5 interviews via zoom (we are now in month 11 of Quarantine) and used my trusty app. Otter.
The main takeaway here was that it backed up the quantitative data gathered, in that you could see a common thread; the rise of the podcast listener.
An interviewee mentioned something that would prove to be very important to defining my MVP.
“I have been listening to podcasts more than ever…I find myself taking notes to share with others or to come back to for reference.”
I was satisfied with the data collected and again it's a 4-day sprint so you must continue to push through. I now wanted to dissect all the data gathered to find patterns and insights.
I started with the Affinity Diagram where I used the post-its feature on Miro to write down the most important data collected from my Qualitative and Quantitative efforts. I then started to place alike categories together. For instance, how often they use Spotify, where they use it, and what they use it for most. I was then able to label the patterns into groups and lastly color-coordinated the post-is to visually enhance the patterns.
Main Takeaway was again the recurring theme of the increase in users listening to podcasts that motivate and educate.
Value Proposition Canvas- Customer Profile
The Value Proposition Canvas consists of two sides; Products & Services and Customer profile. You start with the customer profile side, which is intended to define the customer’s “job to be done” and the associated positive (gains) and negative (pains) experiences, including Functional, Emotional, and Social jobs.
I went into more depth on “Jobs to be done” in my last article, E-Commerce: Helping communities support local businesses, but just in case you haven't had time to read ALL my articles, here’s a quick recap below.
‘Jobs to be done’ as defined by Strategyn proposes that “in order to understand customer needs in a way that makes innovation predictable, companies should stop focusing on the product or the customer and instead focus on the underlying process or “job” the customer is trying to get done.” Meaning, what does a user “hire” Spotify for.
We use the Value Proposition Canvas in its entirety ( I will talk about the Products & Services side later) to ensure the product or service we ultimately provide is a good market fit with the customer and adds value to their experience.
To give you a better idea of a JTBD, I listed below a few examples pertaining to a Spotify user:
- Listens to a podcast to learn about something
- Listens to music & podcasts during their commute to pass the time
- Creates a playlist for a road trip
- Listens to music and/or podcasts at home while cleaning
- Shares playlists/podcasts with friends that need recommendations
And some gains associated with these jobs:
- Motivation, Inspiration, Enlightenment
- Reduces stress and anxiety
- Helps to socialize and entertain
- Sense of pride and belonging
- Sense of empathy and understanding
And some pains associated with these jobs:
- Oversharing content
- Apprehension in sharing out of fear of disapproval
- Frustrated and overwhelmed
I felt confident in the varied amount of JTBDs and set the Canvas aside until I was ready for the Products & Services portion to validate if the feature/s will indeed be a good market fit.
From the data collected and some assumptions, I started by brainstorming things a user would do/think/ and feel while using Spotify. From there I was able to find patterns that translated to phases.
Activity — Listening while: working, exercising, at home, commuting
Discovery — Looking for: new songs, playlists, artists, podcasts
Customizing — Creating: playlists, cover art, editing playlists
Sharing/Social — Sharing playlists, songs, podcasts, listening with friends, group sessions
From the insights gathered in the As-is Map as well as user research, I was able to hone in on a specific user journey in order to uncover specific pain points to address. The journey was, User wants to share a great podcast they have discovered.
I was able to establish 3 pain points that when dissected would become areas of opportunities.
I converted the pain points into more detailed problem statements.
How Might We…
I then converted the Problem Statements into actionable ‘how might we’ statements to help keep me on track and focused on the main pain points.
I was now ready to ideate solutions for the 3 pain points. Using the HMW statements as the framework, I came up with 21 ideas. Brainstorming is not as fun when you are doing it alone. Without the ‘yes, and’ technique, it really is a struggle to pump out ideas in a short period of time. I used the timeboxing technique and gave myself 12 minutes for each HMW statement. There are definitely a few doozies in there which is why I am not zooming in.
In case you don't wake up in a cold sweat from dreaming about the Double Diamond process as I do, you might not know that we are at the Deliver Phase, the second-to-last phase!
I was now ready to dissect my ideas in order to filter down to the Minimum Viable Product!
To do this I started with the MoSCoW method, which allowed me to filter the ideas into must-have, should have, could have, and won't have. I determined that there would be 3 must-haves to form my MVP.
I also applied the Impact and effort technique, which helps emphasize if an idea is impactful with the least amount of effort, which is the ideal situation for an MVP. By the least amount of effort, that could mean costs to build it out and workload for developers. The whole purpose of an MVP after all is to get it made quickly to test faster and more often, improve, rinse and repeat.
I concluded based on data and research collected thus far, that the three Must-Haves I’d focus on would be:
- The ability to cut a clip of a podcast you’d like to share and/or save
- The ability to add a song to multiple playlists at once
- Virtual customizable Bookmark
Value Proposition Canvas- Products & Services
I now needed to ensure what I believed to be must-haves, was a good product-market fit. I came back to the Canvas, focusing on the products and services side to make sure they offered gains and relieved pains or frustrations for the user.
While with more time I most likely would have added some should-haves and could-haves, I needed to concentrate, I only had a few hours left in this sprint.
After concluding that the features would provide value to the user, to back up my findings I used the Jobs story framework as that allows you to focus on context instead of assumptions. First I wanted to get a grasp on what the company’s overall job story is or why the user hires Spotify.
Essentially, Spotify’s story is their mission statement, “to unlock the potential of human creativity — by giving a million creative artists the opportunity to live off their art and billions of fans the opportunity to enjoy and be inspired by it.”
From there I focused on a user’s very specific journey. By seeing a job in a specific context, it allowed me to hone in on a solution that would bring value to the user’s experience.
MINIMUM VIABLE PRODUCT
Yes, we are so very close! I feel like you might have forgotten what my HMWs were, so to recap:
How might we help the user share only certain moments or parts of content?
How might we help the user save their place to be able to revisit or to continue where they left off?
How might we save the user time and reduce unnecessary indecision when sorting new music?
I know you are probably not as excited as I am, but after 4 days of eating and breathing the Design process, I am so excited to have a solution and can get to prototyping an MVP I am actually fond of, so try and get pumped with me.
A feature that allows a user to clip a specific section of a podcast or song in order to share and/or save as well as a virtual customizable bookmark to continue where you left off easily and quickly. Additionally, allowing a user to save a song to various playlists in one same action.
Feature 1: The “Clip” feature will allow the user to select the beginning and end time and with a click of a button will have the specific clipped content saved and ready to share.
Feature 2: The “Virtual Bookmark” feature will allow the user to drop a customizable virtual bookmark at a specific place in order to easily continue where they left off.
Feature 3: The “Multiple Save” feature will allow the user to save a new song in different playlists in one swift action without having to repeat the same action as many times needed.
MVP-Product Market Fit
I focused on the main feature for this challenge and to ensure these features would be a good product-market fit, I researched the company a bit more and found that since 2019 Spotify has invested over $500 million in original podcast content. This was followed by a quick rise in MAU sharing podcasts recommendations. In 2020, Spotify would then launch editorially-curated podcast playlists, to mirror its hugely influential music playlists.
Additionally, I looked back on initial user research both quantitative and qualitative which helped to reinforce that the MVP would add value to the user experience.
- Survey: 77% use Spotify to listen to podcasts
- Interviews: 4 out of 5 listen to podcasts (mainly educational & motivational)
“ I have been listening to podcasts more than ever…I find myself taking notes to share with others or to come back to for reference. ”
By implementing the new feature Spotify would move towards the blue ocean as they become more multifaceted by paying more attention to podcasts and thus welcoming users that might not be music-savvy but enjoy podcasts.
Now I was ready to start prototyping but first I needed to create a flow outline that would help me map out the User’s Happy Path, a scenario featuring no exceptions or error conditions.
The Happy Path: A Spotify subscribed user learning about the new feature than using the feature.
Alright, here we are, hopefully, you made it this far. We are now at our final phase — prototyping and creating our MVP!
We start with a lo-fi prototype, as it is quick, easy, and inexpensive. It allows you to get the product testable quickly for feedback to then improve and test.
Usability Testing — Lo-fi
I submitted my prototype to Maze, a rapid testing platform that allows you to get quick insights and shows you, among other things, how many clicks it took a user to perform the task requested. The overall results were based on 12 user tests and were honestly surprising:
- 0% Mis-click Rate!
- The average duration was 20. 2 seconds
However, there was one particular page, seen above, that when a few of the users were narrating their process, came to a crossroad when asked to go to the podcast section. They said that they almost thought the ‘button’ looked just like regular headline text and not a pathway. However I needed the feature to be native to the app, and therefore no changes of that kind could be made. I was confident that once in hi-fi, it would be more clear and their mental models of Spotify would kick in.
I moved on (quickly) to my Mid-fi prototype…
Usability Testing — Mid-fi
I inserted that puppy right into Maze. Which again, surprisingly, had less successful results than the lo-fi. What I do find is that they tend to get more confused about Maze and the instructions than the actual content.
I wanted to focus on the screens that depicted the two paths of the added feature since the rest is what's native to the app and already typically adopted by the user.
Based on 12 user tests:
Path 1:22% misclick rate
Path 2: 11% misclick rate
I think I definitely need to put more importance into a clear mission instruction on Maze and more clearly brief the tester prior to using Maze as it can be confusing when using on mobile.
But one important takeaway from the results was from one tester who said that she almost didn't see the checkmark that indicated the recording had started and stopped. Again, I wanted to keep it as native to the app as possible, so decided to just increase the size of the checkmark and moved it up towards the progress line.
Now that I needed to move on to the Hi-fi prototype, I built out my Atomic Design Inventory, to assist in creating a consistent design and speed up the process.
I hope you are excited if only to finally finish this article! Please check out my Hi-fi prototype below:
If you even partially thought “ I would likely use that feature and it's pretty intuitive”, that makes me beyond thrilled. I come from a Hospitality background so learning how to design intuitive digital experiences and creating innovative solutions based on user-centered design brings me joy.
Success & Failure Metrics
In order to measure the success of the feature, you need to implement and define success and failure metrics.
- High percentages of sharing “clipped content”
- Increase in both DAU/MAU
- Low Time on Task
- New users
- High churn rate
- Low usage of clip feature
- High misclick rates
- Usability Test the Hi-Fi prototype
- Conduct additional research on any knowledge gaps
- Work on remaining Must Haves, Should haves
Research is King. I cannot imagine doing this process without proper research, you risk basing your whole MVP on only assumptions. I really pushed myself throughout this process to just keep going, be satisfied with what you've gathered so far, and find the insights best you can. I really allowed myself to stick to timeboxing whenever possible. I do need to improve on prioritizing Figma organization. I started strong but as the deadline gained speed I prioritized getting the design done, which of course led to the dreadful Figma Design debt. I think with practice the balance will become easier to keep.
Well, THANK YOU for reading through my design process for the ‘Add a feature’ challenge, I appreciate it and hope you enjoyed the read.
Take care, hope you come back next week.