The beauty of negative churn
It’s been a year since we started monetizing the Literature product at Pratilipi. It took us almost a year to iterate over multiple experiments to figure out what can drive us a steady and scalable stream of revenue. Down the line 12 months, It’s clear now that the Platform-level subscription model is the primary and potential revenue stream for us. We have seen a consistent smiling curve in the repeat purchases. Here are some insights we’ve gained along the way. Please feel free to ask any questions or provide feedback.
Note: For people who are not aware, what does Pratilipi Literature do?
Pratilipi Literature is a UGC storytelling platform for Indian languages. Readers can consume long-form, multi-episodic web series in their preferred languages, self-published by indie authors on the platform. Apart from reading, readers can also engage with the authors by following and messaging them, rating and reviewing their latest episodes, and more.
What’s the platform-level subscription model?
On Pratilipi, Readers read long-form, multi-chapter stories. Less than 5% of the total stories published on the platform contain any paywall. The rest of the stories are completely free.
These paid stories are of two types -
A. Completed stories — these are the popular finished stories published by the authors on the platform
B. Ongoing stories — these are the stories currently being written by the authors on the platform where the latest parts are still arriving at regular time intervals.
The initial ten chapters of these paid stories are free to read. After that reader can choose to proceed with either of the three ways.
- Readers can wait for one more day to unlock the next chapter or
- Readers can pay per chapter or Readers can buy the platform-level subscription to unlock the entire pool of paid stories in one go.
If readers chose not to pay ever then too they can read by waiting one more day to read the next part. In this case, readers get the In-app updates and push notifications after 24 hours that the chapter they were waiting for is unlocked.
Platform-level subscription costs 150 INR a month. And, individual chapters cost 2.5 INR per chapter. We currently have 2.5 lakh active subscriptions.
M12/M1 subscription rate is 114%. Readers who drop out of the subscription program in the first month, start coming back from the second month onwards, and churn eventually becomes negative.
This behavior of negative churn is primarily driven by the very high frequency (~60% of subscribers use the app every single day) and engagement (3.5 hours per day reading time) of subscribers. This leads to very high retention (~85%+ M12 retention), which results in a smiling curve in repurchases.
Let’s understand how this is achieved. I have tried explaining here a user life-cycle from onboarding a user to M12 retained platform subscriber. Please note that in this entire discussion, I have talked about App as a platform for ease of understanding.
Onboarding
In the product, we actively solve the onboarding of new users via continuous experimentation. Once users have installed the app, we optimize for their activation on D0. We define the activation metric as a first-read completion. An activated user has an M1 retention of 53%. Once the activated user starts engaging with the product, our ML recommender system guides them toward a personalized reading experience based on their previous reading preferences. Activated users tend to read ~50 chapters in the M0 and end up following ~4 authors in the M0 as after reading their stories they would like to get more such stories from the author. This brings the community quotient into play.
Once new users have finished reading a free story, recommendations start showing them personalized paid stories of the authors they follow or based on their previous reading preferences.
Activated users usually hit the paywall around 16 times in M0 in their journey of reading these personalized paid stories. Where the first ten chapters are free to read and then to read the next chapter, they can either wait for one more day to unlock it or pay and read the chapter with 2.5 INR or buy a platform-level subscription to unlock all such locked chapters on the platform at once. Once they start reading the paid stories, 54% of these activated users fall into the ‘wait for one more day to unlock the next chapter’ funnel in the M0. Let’s call this funnel ‘early access’.
Now we need this early access funnel to do their first transaction. But they are in the classic dilemma of paying right now or waiting for one more day.
Should they buy coins (make smaller purchases) or directly buy the platform-level subscription (a big purchase)?
I have heard many people on the internet discussing that Bharat users don’t pay, and to be honest, they are not entirely wrong. It’s a fact that nobody wants to pay for something that doesn’t provide enough value. It’s important to understand that these users are making a payment for something they have never experienced spending money on before, and on top of that it’s online! Think about the turmoil you go through before adding another substack newsletter to your ongoing subscriptions. It’s a long process of trust-building on the value front. It’s not on users. The burden is on us to deliver consistent value that is worth the money.
Giving out the very first unlocking experience for free
We wanted these users to try out coins to have their first unlocking experience to continue reading without any friction. We launched a feature called the ‘reading challenge’ where you need to complete the reading streak to earn bonus coins. In a month you can earn up to 25 coins. (you can unlock a maximum of five chapters per month). Out of 54% of M0-activated users who are in the early access funnel, 42% unlock at least one chapter with bonus coins. Now, they are familiar with the value of coins in the system. As a next step, we would want these users to buy coins and continue their reading.
The first transaction
In one of the user surveys, we figured out that 20% of our first-time paying users have never transacted before via smartphone on their own. We got to know that some of these users go to their nearby kirana shops and ask the shopkeepers to purchase on their behalf and then pay them in hard currency. The need for inventing here was very clear to us to gain user trust to transact online on Pratilipi. After a couple of iterations and experiments, we were partially able to figure out what was working for these users who have never transacted before and we launched two features to bridge this penny gap.
1. Unlock the chapter in 1 INR -
As we saw above, we have in-app currency ‘coins’ to unlock the chapters. Users can buy coins that are listed on our storefront in bundles. The minimum denomination bundle is 10 INR to buy 20 coins. 10 INR might be a bigger purchase to start with, for users who have never transacted for content before. We launched a feature where users can directly transact without purchasing coins and unlock a particular chapter for 1 INR. Just one click via UPI. Daily around 600 users started transacting via this feature. These are the users who were in the early access funnel for a long time.
This feature is only available to users who have never transacted before on their very first purchase.
2. Ask your friends/family to pay -
We started giving out the WhatsApp shareable link for payment on the checkout screen. 100 users are using this link to do payments daily. We have only opened up this feature recently and to the never transacting users. We are planning to open it to a broader audience soon.
Coin buying
By now they already have used the bonus coins and transacted once to unlock the chapters. 10% of these users who do 1 INR transaction, convert into coin buyers in the next 7 days. They start with a lower denomination like 20 coins in 10 INR to keep unlocking the next chapters. They gradually shift to higher denominations. 70% of the daily coin buyers are repeated coin buyers with an average value of 50 INR (100 coins) denomination.
Coin buyers read 4 paid stories simultaneously in a week and hit the paywall 48 times. Meanwhile, a point comes where transacting on each chapter doesn’t make any more sense and they finally get into the platform-level subscription.
21% of these first-time coin buyers convert to platform-level subscribers within 30 days.
Problems with Discounts
Offering initial discount prices for first-time platform-level subscriptions can be a lucrative way to attract users to the subscription cycle. However, we have also learned the hard way that discounts can be a slippery slope. What starts as occasional discounts for festivals or other special occasions can quickly turn into a habit of using cohort-based anchoring to convert a specific funnel. The flood of subscription numbers due to discounts can lead to a vanity trap, where we become focused on numbers rather than the true value. Unfortunately, using heavy discounts to drive subscriptions can be a dangerous game. While it may initially attract price-sensitive customers, it can eventually lead to increased churn and a decline in organic conversions at higher price points. This is detrimental to the business in two ways.
- Users start expecting lower price points and will wait for the discounts. The same users would have otherwise paid at the non-discounted price points.
- If you do not quickly correct the price-sensitive subscribers’ recurring subscription rate then eventually you will take the hit on the annualized revenue per paying user.
Path to repeat subscriptions
At the beginning of this post, we talked about the community quotient. Let’s understand how it plays a major role in achieving a smiling curve in repeat subscription behavior. As we know, users are simultaneously reading multiple paid stories at the time of buying a platform-level subscription for the very first time. Users could be reading either of the two types of paid stories before getting into the subscription program. It’s very hard to attribute the exact purchase trigger but their buying intent might be coming from reading ‘completed’ paid stories or ‘ongoing’ paid stories or reading both. Subscribers chart different journeys after buying the platform-level subscription for the first time, depending on their reading behavior at the time of purchase.
Let’s understand one such journey where subscribers are reading ‘completed’ paid stories at the time of buying the platform-level subscription for the first time.
These M0 subscribers are potential churn for M1. They bought the subscription to read certain ‘completed’ stories and once they binge-read them then there is a good chance that they go out of the loop of paid story reading behavior. However, these are engaged users and there are plenty of free stories available on the platform, they will not churn out of the platform altogether. Just that immediate buy intent would dilute. Our effort here is to primarily make sure that in W1 to W4 of their M0, they are engaged with the paid stories throughout.
For subscribers who are engaged to paid stories in their w4, ~80% of them end up continuing their subscription in M1.
We achieve W4 engagement to paid stories by -
- serving super personalized ‘completed’ stories in their recommendation
- increasing the pool of quality stories under the paid program for the personalization to run better
- Hook these users to ‘ongoing’ paid stories in their W1 itself.
I would like to emphasize the third point here as that’s where the community quotient plays a major role and the beauty of cross-platform network effects directly reflects in the repeat buying behavior. As we saw earlier, ‘ongoing’ stories are by nature not finished. Authors are dropping the latest chapters at regular time intervals. And, once the subscribers have started reading the ‘ongoing’ paid stories in M0, then they will wait for the latest chapters to arrive. So, W4 engagement to paid stories in M0 goes significantly up for these M0 subscribers, and in turn, their M1 is also kind of taken care of.
What happens to the M0 subscribers who churn?
The aftermath is more nuanced. Get a cup of tea. :)
Before we look into what happens to M0 churned subscribers, I would like to tell more about how a new user becomes a part of a community on the platform. When a user follows an author, they begin to participate in other features and product loops as well. This includes reviews and comments from the fan community, as well as direct updates from writers to the fans. As a result, users are not just returning for the stories but are also becoming a part of a like-minded community of readers. There is more to it if you are a platform-level subscriber. when you take a platform-level subscription you get exclusive access to authors’ fan community chatrooms as well. In Chatrooms, like-minded readers primarily talk about the story plot, characters, latest chapter updates, etc and the author directly participates in these conversations. When you go out of the subscription ecosystem, you lose access to these chatrooms.
Because we actively started engaging M0 churned subscribers with ‘ongoing’ paid stories from their W1 of the M0 or even before that, they are still reading these stories and hence falling under the early access funnel. However, the buying intent isn’t strong enough as yet. The M0 churned subscriber is primarily following the author to get push notifications and in-app updates regarding the latest chapters coming out of early access. Meanwhile, the same author, whose story M0 churned subscribers are hooked to, starts another fresh new paid story. Now the FOMO starts triggering to the M0 churned subscribers. As he/she starts getting the latest updates regarding what’s happening in the author’s fan community without being a direct audience. He/she will be naturally inclined to start reading this new story as there is enough hullabaloo in the fan community about this new story. In the reviews and comments, the fan community is discussing what has happened in the latest chapters. And, that eventually creates a massive FOMO to be part of the chatrooms once again and get all the latest chapter updates on time. Now assume all this happening when you are following more than five such authors and you are part of their fan communities. A point comes where the frustration is not worth and churned M0 subscriber restarts the subscription again.
Subscribers who are engaged in ‘ongoing’ paid stories and are part of the above-described community, tend to have 70%+ M10 retention. They gradually upgrade to yearly plans.
Along with the above community-led network effects, robust personalization also helps in keeping the churned or existing subscribers in the paid story reading loop. In addition to value retention, improving the recurring payment experience has also proven good for us. The majority of our paid users, around 86%, transact via UPI, which prevented us from using subscription payment products until recently. To address this, we partnered with PhonePe as a UPI auto-pay partner, which has further increased our re-subscription rate. While it is still early to draw conclusions, we have seen a resubscription rate of approximately +75% among subscribers who use PhonePe as a payment method.
How do the supply-side dynamics work in the subscription play?
Everything I discussed above becomes incredibly significant when the cost of content production is nearly zero. This enables us to operate at gross margins of over 60%. When user-generated content (UGC) begins to scale and operate as an autonomous factory, it creates a multiplier effect that enhances everything. However, the workings of the supply side of things in this model is a separate story, which I will cover next time. :)