Spotify has a 1-month trial membership. We noticed that users were not converting to paid membership after the free trial. From a product standpoint, what can you do to improve user engagement and subsequent customer retention?

Question

Spotify has a 1-month trial membership. We noticed that users were not converting to paid membership after the free trial. From a product standpoint, what can you do to improve user engagement and subsequent customer retention?

Asked at Spotify over a year ago

8.7k views

Product Improvement category

Asked at

8.7k views

How to answer Product Improvement questions

Interview Guide

Top Spotify interview questions

What is your favorite product? Why?89 answers | 263k views
How would you improve YouTube?29 answers | 81.3k views
Design a library for the future.25 answers | 23.5k views
See Spotify PM Interview Questions

Top Product Improvement interview questions

How would you improve Google Maps?53 answers | 228k views
How can you improve Facebook Stories?22 answers | 45.5k views
How would you improve Facebook Birthdays?21 answers | 25.8k views
See Product Improvement PM Interview Questions

Invite members
Invite by email

Add another

Prateek Sachdeva · Answer 1 · 2023-10-08T03:05:19+0000

Clarifying questions:

Are we specifically talking about users on free trial not converting to paid membership? (Yes)
What features of Spotify are available during the free trial period? (ad-free music, playing while offline, playing on multiple devices, downloading, blocking certain type of music)
There are multiple plans of Spotify. For which plan specifically are we observing this? Across all plans
Any specifc type of user that is churning within the free trial period? (users with low engagemnent on the app during the free trial) (casual users)
Are users able to access downloaded content after the premium plan expires? No
When are the users charged for the premium account? First day of every month.
Are we experiencing this for any specific geographic region? India
Do we have any technical / budget constraints? No

Goal:

Spotify is a music/ audio-content streaming platform that brings users personalized, curated content. It’s premium version offers ad-free music. In the non-premium version users experience ads. Spotify’s two streams of revenue are: ad-driven & subscription driven.

Our goal is to boost user engagement for users on free trial so as to ensure that users are retained after a month.

Users:

What kind of users sign up for free trial plans:

Students who are operating on a budget and are music lovers (can’t expand their budget: India is a price sensitive market)
People who workout and like continuity in streaming of music
People who listen to music / podcasts while driving (are on road for a long time)
People with busy work schedules who consume a lot of audio content passively and don’t have time for video content. (won’t mind ads, Spotify Premium’s USP)

Out of these cohorts let’s identify users who would dislike ads, like to download music and have the ability to play it on multiple devices

Priortising cohort 2 & 3.

Painpoints:

find it a hassle to search for songs and play while driving/ working out (hands-free mode)
are confused about what kind of music / podcast they are in a mood for
find it difficult to connect with car speakers
are not able to discover content that they like
sometimes they don’t want to listen to all songs in an album.

Features:

Spotify hands-free mode: (solves for 1 & 4) P1

AI-powered voice assistant that would auto detect when a user is working out / driving and will be switched on. This voice assistant would assist the user to navigate spotify without them having to access their phone/ smart watch. The assitant can find new songs, play a certain song from an album

ML based album curator: P2

Users who are on a free trial plan would have their moods monitored closely. This algorithm would study user behaviour basis on the type of songs that they are listening to, when was this music released, observe artists, lyrics (sentimental analysis using NLP), if they are wearing any smart-devices then it would connect with that as well. Basis on these factors it would classify each music streaming session basis certain moods like sad, romantic, party, nostalgia, workout mode etc. These mood albums would be auto populated. By the time the free trial month ends, there will be a couple of albums. The user would not have to search for songs and can simply ask the AI assistant to play a certain mood album. This can be replicated for podcasts as well.

Spotify music/ podcast finder: P3

Many a times users come across a reel/ short form content that is from a podcast/ music video etc and the users are not aware of the audio. It’s very hard to find it on Google as well.

We will build a finder where the user can find the exact content that they are looking for basis:

uploading a link to the content (tiktok/ Instagram etc.)
name of the artist (not mandatory)
typing in lyrics / content description
Product integrations with Instagram & TikTok (when a certain reel is playing, there will be a spotify icon right next to it, when the user will click on the ico, he/she will be redirected to Spotify where they would be able to find the content)
Voice recording (a user can also record a tune/ song and upload it, Spotify would analyse the lyrics and tune and then find the content on its database)

Listen to what your friends are listening: (P4)

Integrate Spotify with your Instagram account. Spotify would look at you connections and curate a list of 50 songs & podcasts every week which would have suggestions of what your connections are listening to. This list would be updated every week. We would be including the top most frequently listened to songs.

Success Metrics:

D30, D60, D90 retention rates of users on free trial (North star)
Churn rate of free trial users
Engagement rate of users on free trial (are casual users converting to core?)
No. of premium users using hands-free mode & average session length of hands-free mode
Average spotify session time & DAU
Number of searches on Spotify content finder (basis what are the users searching segmentation)

Chetan Channa · Answer 2 · 2020-05-31T08:00:13+0000

Let me share my thoughts on answering this Spotify product improvement question.

Assumptions

We would only be focussing on users who we acquire on Free trial
We will assume that the user journey where the user gets acquired on paid directly is already there for comparison.
We will do the below analysis across all platforms (desktop, mobile apps etc)

Let’s look at the typical user journey

(1)User comes to the product -> (2)Is exposed to Free Trial option -> (3)Gets a Free Trial -> (4)Engages with the product in the same session ->(5) Returns and engages with the product ->(6) Buys Subscription at the end of free trial -> (7)Engages with the product

The problem statement we have is

Is to improve Conversion Rate between 3 & 6 (this will cover both engagement & free to paid conversions )

Lets look at individual steps between 3&6

Step-1 : (3)Gets a Free Trial -> (4)Engages with the product in the same session

The underlying assumption here is that unless user experiences/likes the product, the chances are very slim that he/she will convert. Moreover,

For this analysis, I will focus on 3 metrics :-

% if users listening to at least one audio after availing free trial- This number should be close to 100%. Since the user has availed the free trial, he has at-least some motivation to use the product. If this number is not close to 100% then, we can have 2 hypothesis
1. Technical issue with the product for those users which led them to leave the product without listening to even one audio. If I find an issue, I will fix it & reach out to impacted users by push notifications/emails or any other available mechanism.
2. User didn’t find any audio/content interesting enough to listen. It doesn’t make sense to target these users and hence I would do a further analysis on these users to figure out the segment which we should not target. Analysis can be done basis :-
  1. Source of Acquisition - A particular acquisition source/campaign is not giving good results.
  2. Country of Acquisition
  3. Other segmentation criteria like gender, age etc. These are hard to get but this can be achieved by doing a survey of such users either online or offline.
Number of audios initiated per user - I would keep a track of this metric and make sure that the count is in the same range as of users who directly paid. This is basically the sampling rate and if the sampling rate is not good enough and chances are slim that the users will convert. To improve sampling rate, I would do following things
1. Experiment & improve first time recommendations
2. Showcase any audio/content which is exclusive to Spotify
3. Make discovery super-easy
4. Make sure that search is working fine and users are able to find the music they like
Total minutes of audio listened by users - I would keep a track of this metric and make sure that the count is in the same range as of users who directly paid.

Step-2: (4)Engages with the product in the same session ->(5) Returns and engages with the product

After the first session, we need to make sure that user comes back again. in most of the products, this is the major drop and if the product can ensure that the user comes back the 2nd time, the probability of user sticking for longer improves significantly. Following things can be done

Run re-targeting campaigns for free trial users after their first visit.
Send emailers/push notifications with exclusive content and/or recommendations for the user.
If the user has signed in via Fb, then post user’s friend’s activity on Fb wall.
Do a survey to figure out why they don’t want to return.

Step-3: 5) Returns and engages with the product ->(6) Buys Subscription at the end of free trial

Now there would be substantially large number of users who engaged with the product but didn’t buy subscription. Ideally I would do a funnel analysis for this step and figure out drop at various steps. A sample funnel would be (basis common payment journeys across products)

(1)Free trial ends -> (2)User Initiating subscription & gets to see the payment options/plans -> (3)User chooses a plan & Initiates payment -> (4)Successfully completes payment

Following can be the hypothesis for drops at each step :-

Not enough value in the subscription (For drop between 1->2)
1. Create more value in subscription and/or create more pain points in Free trial. For eg introduce more ads, limit number of audios in free trial, limit number of minutes a song can be played etc
Difficulty in initiating subscription (For drop between 1->2)
1. Do UI/UX improvement to make sure it is super easy to subscribe.
Subscription was too costly (For drop between 2->3)
1. Introduce cheaper plans. (There can be many innovative ways of making plans like for eg, you can make a plan for songs for one artist etc.)
Didn’t find a suitable payment partner (For drop between 2->3)
1. Apart from Credit Card, introduce more payment partners like wallets etc.
Technical issue during subscription or payment Failed (For drop between 3->4)
1. Fix the technical issues and re-target such users

So by looking at drops at various steps of ser journeys and fixing those drops , we should be able to improve retention as well as free to paid conversions.

Some other ideas

Mobile apps tend to have higher retention and engagement than mobile site. Also if the user has access to products on more platform, probability of retention and engagement will become higher. This means that I will introduce flows and features which make user download and use product across platforms. For eg for desktop/mobile site user, introduction of feature to download app would help, for mobile app user, letting the user know that there is desktop site would help.
Limited Coupons and discounts do help. We can introduce discounts and coupons for users dropping off at various steps, but we need to ensure that we don’t degrade the value of the product by deep discounting.
Last but bot the least, I would definitely do a deeper analysis of the funnel via source/campaign of acquisition. Generally direct/organic users would be better and will make sure that mix of acquisition medium is appropriate basis our goals.

Ark · Answer 3 · 2020-05-29T20:05:20+0000

I would go about this in this particular flow :- divide this into external and internal factors (both explained below). After analysing the factors will we go for solutioning.

Internal factors :

These are the factors which are under the purview of Spotify and altering them can change customer behavior. There are users that are not converting and users that are converting. A comparison needs to be made between the two to check what is working for one and what is not working for the other. This will bring out the 'motivation' which made one user convert. Users who did not convert should be analysed across the following metrics -

Is it happening across mobile app only users or desktop app only users or both ?

Are they music listeners or podcast listeners ?

Do they belong to a particular age group ?

Do they belong to a certain geography or are they scattered across regions ?

What was the average listening time of the users?

Is there any particular genre to which these users were listening to?

On an average how many songs were liked and how many artists/playlists/albums were added ?

What was the average number of times they skipped a song before 30secs during that 1 month period ?

What was the average income level of the users?

What are the professions to which the users belonged to?

Now compare the above metrics with the users who had converted and notice any difference. The next level of analysis should come from customer interviews itself - interview both non-converted customers as well as converted customers and ask about their motivations -

Why did one start a free trial membership - was it to block ads? was it to explore more recommendations ? etc

Why did one subscribe OR not subscribe from a free trial ? were the recommendations relevant/not relevant

Final step would be to club these 2 levels of analysis and search for any connections.

External Factors

These are factors over which Spotify may not have direct control. For me these will mostly be across competitors of Spotify - is there a new rival app ? Are they providing better pricing ?

Solutioning

Solutioning is based completely upon the answers to the above questions. To ennumerate a few -

If the issue is across a particular age group - dig deeper to find out what is the most preferred genre. Improve the recommendations & content for that group. Notifying users of any new sing added by their favourite artist/ or any new song added to their favourite genre might be a good idea

If users are skipping more songs - again this might be a content issue. The recommendations are not relevant to the user

If its happening across a particular income group (for eg students) - Dynamic pricing of plans.

.

Nishant Kumar · Answer 4 · 2024-05-26T07:44:16+0000

CQ:

There are multiple paid plans, which plan are we interested in -> across all plans
Any sensitivity to platform -> web over mobile
What features are available in the trial period -> no ads, unlimited songs, others
Geography -> not sensitive
Sudden or gradual -> happening gradually since a quarter
Do we see any drop to other metrics like -> new acquisition, % conversion to free trial, user engagement during free trial -> make your assumptions and proceed

Hypothesis:

% conversion to free trial has dropped:
1. Availability -> introduce new discovery points in the user journey
2. Likeliness ->
  1. reduce the number of steps required to enrol
  2. UI enhancements; Talk more about benefits
3. Awareness -> customer outreach via multiple channels to bring them to the top of the funnel
User engagement during free trial has dropped:
1. Perform user research/competitive analysis for various user segments and understand the user needs -> build the prioritised asks into the product
Process of conversion to premium is buggy -> Get it QAed and resolve the errors

Metrics for improvement:

% users opting for free trial
% free trial users who converted to premium
% DAU/WAU users who are on free trial

Lokesh Kumar R · Answer 5 · 2021-04-28T17:30:49+0000

Clarifications:

Spotify has its customers around the world. Are there any particular pattern from where users are showing such a behaviour? Yes, These kind of patterns are more prevalent in the Asian Countries, Espescially India.

From the PS, it is pretty obvious that our goal is to increase the users in Premium categories or in short, improve User Activation to Premium Plans. Am I missing out anything? No, You've got it right.

Spotify is offering different Premium plans. Is there a particular plan that is facing such an issue? No, the issue is pretty much present in all the three categories.

With the above clarifications, I would like to jump on to improve the product espescially for Mobile apps to solve the particular issue. The reason behind choosing mobiles are Spotify has its name for listening to music/ podcasts on the go and Mobiles are the device that people use on the go compared to other devices.

First I would like to share my understanding of the product to make sure we are in the same page. Next, I would discuss the User Categories, Prioritize one from them and then List down their pain points. Later, I would prioritize the pain points and then list out the solutions. As the final step I would like to prioritize the solutions and do a deep dive over there.

Background:

Spotify is a service provider providing users with millions of songs and podcasts through medium like Web app, Phone App (Both iOS and Android). The company's vision statement is as follows-

We envision a cultural platform where professional creators can break free of their medium’s constraints and where everyone can enjoy an immersive artistic experience that enables us to empathize with each other and to feel part of a greater whole.

The strength of the firm rests in their innovative way of personalizing the Music Experience with the users.

With this understanding, Next I'll jump on to the user categories.

User Categories:

Since the goal is to activate users to premium services, I would like to categorize the users (who are still free users) depending on their usage of the product.

Power Users- Users who listen songs for hours together in the app per day.
Medium Users-Users who listen to songs for hours in the app by the end of the week
Occasional Uses- Users who rarely listen now and then randomly.

Of these users, we shall prioritize the first type, i.e, Power Users, as they are more interested in the app that they would need just a small push to activate them into premium users compared to others.

One way to increase user conversion to premium service is by increasing the value proposed. So let's analyse the pain points of the users and come up with a solution for premium mode. With this, lets jump on to discuss the pain points.

Pain Points:

Enjoy streaming services without forced breaks like ads
Finding the right song for the mood/situation
Finding a new song matching your taste from the millions of songs are getting more more and difficult.

Of all the points above, I would like to focus on the 2nd and 3rd point as these will add huge value to spotify in easening the process of streaming and alligns more with the vision. While the 1st one though a big problem People tend to adjust with this as this issue is something that is common to every other app.

Solutions:

Spotify can come up with a new set of playlist every week with new songs that a user hasn't listened to and that are in sync with their listening that week from different languages suitable to their location.
Adding a new search feature that enables user to search for songs by humming the tune of the songs or by searching a song through the movie's cast and not the just the singers.

Solutions	Reach	Complexity	Impact
Solution-1	Medium	Easy	Medium
Solution-2	High	Medium	High

The reason behind solution 1 having a low priority to the 2nd solution is that often new set of songs comes to the knowledge our chosen user through word of mouth. So they don't depend on just an app for this.

While the second, is a major problem as often people forget the name of the song and remember the tune or the hero and heroine. To get the name of the song the usual method is to use google, youtube or other similar sources to get the name. This causes deviation in the user journey which is not good for us. Thus having such an option will help people find the song a lot easier than the usual text mode.

Features:

To implement the following, first of all every song details has to be updated with the Cast of that movie. Next, The search algorithm should be updated to use the cast names as a factor.

Spotify can use ML models to identify the humming of a user and list the most likely song. One important thing to keep in mind is that not every user is a good singer. So the similarity checker's boundary condition should be kept a bit low than usual.

In short, this is more of an upgradation to the search feature present in Spotify espescially for Premium users.

Success Metrics:

# of searches made using cast or humming per the usual search- this shows how frequent the new functionality is used.
# of songs played through a search of tune-this shows how successful are we in predicting the right song from the tune.

Summary

The goal was to increase the users converting to premium. By analysing the user pain points and solving them for the premium users will help us increase the value for the premium version. With this in the mind a problem was picked and then came up with the solution of adding extra functionality to search so that users can find the song with the cast details or the tune.

productmanager1644 · Answer 6 · 2021-02-05T05:02:08+0000

CLARIFY:
1. Are we focused on a specific market? You choose.
2. Is there a specific demographic I should be aware of? You choose.
3. Is it OK to assume the metrics are working OK re. conversion etc.? Yes
4. Is it also OK to assume there are no technical bugs that are preventing the users from make the conversion? Yes
5. Is there a specific type of Spotify premium membership I need to focus on? You choose.
PROBLEM: Spotify users are offered 1 month premium member but are not converting after the month is complete to paid users. There are a number of free 1 month Spotify premium memberships. (I am skipping the Premium Individual account because it is listed as 2 months free on my Spotify for now.) For the context of the problem, I'd like to focus on 1 type of premium membership. Here, I am going to assume that the free account is the Premium Student account.
1. Premium Student: Includes access to Hulu and Showtime with ad fere music, unlimited skips, etc.
2. Premium Duo: Provides premium access for 2 accounts. Same ad free music, unlimited skips, on demand playback, etc.
3. Premium Family: Up to 6 premium kids accounts. Contains Spotify Kids app, block explicit music, family mixes, etc.
PROBING QUESTIONS: I'd like to better understand the problem / background context. There are internal and external factors to consider.
1. INTERNAL FACTORS:
  1. User:
    1. Who were the initial target users for the 1 month subscription? Age range? Region? Were these targetted users the correct ones?
    2. Do they lean more towards on type of device: Android v iPhone?
    3. How close to graduation were students when they activated the account? (If they were about to graduate, they may not think they qualify for the student account anymore.)
    4. How long had the user used Spotify before activating Premium?
  2. Turning on Premium:
    1. When did users turn on Premium? What day of the week / time / year? (Ex. perhaps they were playing music at a party and wanted to skip ads for the one party.)
    2. Were they actively marketed to re. Premium / its benefits or did they just turn it on in the app?
    3. Was there a chance the user accidentally turned on premium without realizing what it was (i.e. they would have to pay for it later)?
  3. Engagement: For each engagement metric, I'd like to understand how the user engagement changed before v. after Premium was activated, assuming there is enough data on their use of Spotify before Premium.
    1. How often did the user engage with Spotify before v. after Premium (i.e. logins / month)?
    2. How many songs on average do they listen to each session? How long were the sessions?
    3. How often did they skip songs? Did they max skips before Premium was activated?
    4. How often did they explore new music?
    5. Did they often skip suggested / new songs Spotify would like? (Perhaps the data science model isn't accurate enough.)
    6. How many likes did they make on songs and artists? How many artists did they follow?
    7. What other aspects of the platform are they using: playlists, podcasts, etc.?
    8. Did Spotify send them enough emails during the month to remind them about the Spotify benefits / deadline to activate a full Premium membership?
  4. UI:
    1. Did users often scroll through the app for a while before playing a song? (Perhaps the layout is confusing.)
    2. Are there heat maps available to see where the user engages most.
  5. Cost:
    1. What is the user's monthly income? What is the average spend they're willing to / can spend on subscriptions?
  6. App Bundling:
    1. Did students engage in the other apps as part of their bundle? Same engagement metrics re. music I'd like to apply (where applicable to video streaming services - ex. number of shows watched, length of time spent on platforms, number of logins / month, etc..)
2. EXTERNAL FACTORS:
  1. Competitors:
    1. Was there a new competitor app that entered the space that took away from conversions?
    2. Where else do student users listen to music: Apple, Amazon, Tidel, Youtube, Soundcloud etc.
    3. How do competitors for the target group compare re. music offerings (songs / artists / genres), data science models to predict user preferences, ability to customize the music experience and cost?
    4. Do competitors offer a better package deal re. partnerships for example for students? Perhaps students would prefer a partnership with Disney+ for example.
HYPOTHESIS: I hypothesize students are not converting from the 1 month free trial membership to the Spotify Student Premium account because they do not see enough value in the Spotify Student Premium membership and can get sufficient benefit in the free account or on a competitor app.
SUGGESTION ACTIONS:
1. Analyze the existing target data. Find answers to questions above.
2. Conduct customer research to better understand the target customer:
  1. What do they like / don't like in their music streaming service? What do they most value re. music streaming?
  2. What kind of music do they most like to listen to?
  3. What are deal breakers v. nices to have for them? (Perhaps the ad lengths are tolerable.)
  4. What other subscription services do they most use / want? Are Hulu / Showtime the right partners?
3. Implement changes based on analysis. For example, if I find that students prefer to listen to more mash ups / mixes / live shows, I'd work to bring more mixes onto the platform. Spotify Student accounts could have access to live DJ sets or taped DJ sets only available under the student account.