Poesie Annual Update: Experiences Running a Paid iPhone App (2021)
This post continues the tradition of publishing an annual report detailing the year’s work and learnings from Poesie, my daily poetry app. This report will include some basic stats on the finances and growth of the app, along with a brief discussion around the paid app model. I want it to be a useful resource for anyone building their own apps, so there are lots of charts and data points. Feel free to ask any questions!
Introduction: App(s) Overview
Poesie is a daily poetry app that I built and launched a bit over three years ago at the start of 2018. After spending the next year working on other apps, I returned to focus on Poesie around June 2019. The app was retaining well and growing, and I made it a paid app one year later in May 2020. A highlight of the past year was being featured as the App of the Day in September 2020.
My hope with Poesie is to: (1) offer poetry lovers around the world a service for discovering and engaging with poetry every day, (2) help people feel like they are part of a global community of poetry lovers, and (3) provide a source of income for myself (app developer) and for participating poets (content providers).
In June 2021, I also launched the app Scholar, a daily trivia app. Like Poesie, this is a daily, educational, community-oriented app that offers three short trivia questions each morning, themed around some topic in history or the humanities. I’ve always had a dream to run a suite of apps that help me develop and share my passions with other people. More on this in the final section.
Core Stats: Revenue, Margins, Growth
The past year has been focused on building up Poesie’s subscribed user base. As of July 31, 2021, Poesie has around 1,400 paid subscribers. It feels great to know that this is possible — I never imagined two years ago that this poetry app could become a self-sustaining service that is even able to send small payments to participating poets.
Revenue: I offer a monthly subscription ($1.99/mo) or annual option ($14.99/mo). In our first year as a paid app (Aug 2020 to July 2021), Poesie sold $28k of subscriptions, bringing revenue to about $2,200 per month as of June 2021. Subscriptions are sold directly from within the app, usually upon download or at the end of a short initial preview.
Margins: The main costs associated with the app include: 15% of revenue to Apple per their subscriptions agreement, 15% of revenue dedicated to royalties per my current contract with poets, and about 15% of revenue for operations (variable server cost, fixed corporate fees, a few other services). That leaves about 50% of revenue for salaries or additional investment into the business.
Growth: With a subscription paywall, it does make more sense to track subscribed MAU instead of overall MAU. However, as of July 2021, I moved to a freemium model —hoping to trade back some of those subscriptions for organic growth (and conversion down the line). So while I build out the freemium experience, I track both subscribed and subscribed users. The total user base is currently at around 3,500 MAU.
Behind the Scenes: Conversion and Retention Rates
Growth and revenue are a function of new users, retention, and purchase conversion rates. Here are a few of these important funnel metrics for comparison against your own apps.
Purchase Rate: Chart #3 shows the conversion rate from first use to subscription purchase. For now, around 5–6% of installs convert to purchase. This rate varies with the source and quantity of new users. For example, through April-May 2020, Poesie was ranked a bit lower in the App Store and new users likely had more intention to purchase. Note that July and August of 2021 are a bit undercounted as some users do convert in the months after install. Subscription rates will likely experience some additional decline (seen during experimentation) as Poesie shifts to freemium model.
Purchase retention: Chart #4 shows what the subscription retention looks like for subscribers on the monthly plan. This is most useful for understanding the net revenue from subscribers. Based on these values, someone who signs up on the monthly plan brings about $9 in the first year plus hopefully an additional $3 per year afterwards, i.e. $12 in two years. Not pictured are the annual subscribers, who (so far) retain about 40% their first year, i.e. $20 in two years. Since about 33% of purchases opt for annual, a new user brings about $15 in revenue over two years at these rates.
Install Retention: Chart #5 shows the drop-off in new users retention after the introduction of the paywall. From a 20–25% next-month retention, Poesie fell to about 10% next-month retention. (Almost all of this is captured in the first few days drop-off.) As Poesie shifts to a freemium model, I hope the retention will climb back a bit and create a balanced user base of subscriber and non-subscribers.
Discussion: Paid vs. Freemium Models
The decision between paid, freemium, or ad-based models depends on my aspirations for the app itself, revenue goals, and the user experience.
Paid: Much of the data shown above is from the paid model, which I kept for a year. I likely would have continued ahead with this approach if I were able to grow quickly with advertising — which means achieving profitability on (roughly) a per-unit ad basis. Given our conversion rates above (5% conversion w/ $15 near-term value), this basically means a download should cost at most $0.75-$0.90. Last I tried, Facebook ads ended up costing around $0.90-$1.15 per download and around $30/subscription. With additional work on the ad copy, targeting, and ongoing work around in-app conversion, I don’t think it is impossible to get close to per-unit profitability. If I was working with a team, this might be worth iterating around. But for now, I shifted focus to a freemium model.
Freemium: The freemium approach aims to grow overall MAU while converting users to subscribers over time. My experimentation confirmed that I earned back retention at the cost of some purchases, and I also saw more app sharing from new users — that was promising. There is still time and work needed to see how the subscriber base will evolve in this model. I do feel happy to give a useful product to people for free, though I would likely choose the paid model if it very clearly paved the way to reaching more people and building the self-sustaining community.
Ads: I haven’t explored ads yet. The idea here would be to bring in some small ad revenue while encouraging people to subscribe for a clean experience. Since I am explicitly building a product that is geared towards 5-minutes per day (not huge amounts of time spent), it might be a challenge to earn ad revenue effectively. With additional team members, I wouldn’t be averse to trying it.
For now, I will let the freemium model play out — the lowest-touch of the three options. This discussion demonstrates some of the difficulties working alone and without much financial investment. Given a larger team, I likely would have iterated more deeply on getting closer to per-unit profitably (in order to be able to run ads for growth) and may have also experimented with in-app ads as a revenue source and subscription incentive. Working alone, I am limited by my personal preferences/ability in terms of time, product direction, and motivations.
Upcoming Plans: Trivia App, Poesie Growth, Operational Sustainability
Near-term efforts and goals in the next few months include:
- Scholar Trivia App: Product work and initial growth work. Retention and organic sharing are not quite where I want them to be — a few more points of engagement across the app may be helpful. We may also be not reaching the target user base effectively, and I may invest in a small batch of ads to try and reach a higher, stable app store ranking. In the long run, I would like to see if it is possible to even earn some income from this app — perhaps with ads and a subscription option.
- Poesie Poetry App: Ongoing growth work. There are a few features to build that may improve the freemium experience; a stable, upwards trend for both overall MAU and subscribed users, albeit slower than before, would be great. I would also like to try out some alternative marketing approaches: I have heard stories from users about (a) literature teachers sharing the app with their classes, and (b) readers sharing the app among their poetry groups. Perhaps offering bulk discounts to educational institutions like libraries or schools may help us begin reaching some of these readers. I did reach out to some libraries a few years back without much interest, but being able to point to our recent App of the Day feature and partnerships with contemporary publishers may help us revisit these conversations.
Poesie is currently at around $25k-30k revenue per year. A total goal of $100k revenue per year from these small apps seems achievable, including a combination of continued growth from Poesie and possibly opening up a revenue stream from Scholar. That would feel like a great source of income that could fund additional apps or hire people to the team. I’ve always dreamed of having a suite of apps focused on helping people pursue educational, interactive, and community-focused hobbies, so getting Scholar into a good place (similar to where Poesie was a few years ago) feels important to me. After that, I may either build a third app or maintain these apps operationally while shifting focus to another type of project.