350 Million Comic Pages Read Later — ComicFlow’s 2014 Metrics & Financials
To follow up on the recent trend from William Wilkinson with Manual, Marco Arment with Overcast, Jared Sinclair with Unread, Bogdan Popescu with Dash, but also our massive financial & intelligence data dump from the Everpix startup early last year, this post reveals the 2014 financials along with some of the usage metrics for my ComicFlow app for iPad.
The Birth of ComicFlow
Soon after the original iPad was released, it was obvious it would become a great device for reading comics (or “illustrated novels” as we say in French). I tried a few comic reading apps at the time, but they all had significant issues: poor user experience, general slowness, bugs, crashes, and were generally incapable of handling large comic collections (say more than afew hundred files).
As the saying goes, if you want something done right, do it yourself. Soon, ComicFlow was born: it had a very simple and focused user interface, was fast, stable and offered a built-in web server to easily import comic files “over-the-air” without using iTunes.
I figured this app might appeal to other people and decided to release it for free on the App Store mid-December 2010.
ComicFlow Today
Now 4 years later at version 1.4.1, ComicFlow is used around the world to read more than 1 million comic pages per day (the average user reads 30 pages in one session).
It boasts a stable 4.5 stars average rating over 7,000 ratings worldwide and is essentially crash-free (Crashlytics reports a 100% crash-free users rate). ComicFlow is even still compatible with the original iPad running iOS 5.1.
Monetizing ComicFlow
Late December 2012, I rewrote the built-in web server from scratch. It had originally been implemented on top of a third-party code base which was showing its age and was barely maintained anymore. This was starting to affect the user experience as the web server was not as reliable as it should have been.
The result of this work was GCDWebServer, a modern and reliable HTTP web server for iOS and OS X, designed with a clean and extensible architecture based on Grand Central Dispatch. GCDWebServer is also open-source and now quite popular on GitHub (it’s used in many apps and by companies like LG, Apple or the Apache Software Foundation).
Rewriting the web server component was the first significant work on ComicFlow 2 years after its launch (probably a month of part-time work but I don’t remember precisely). I figured that since the app had a decent user base and in-app purchases were now available, why not try to charge for this feature? To soften the “blow” of making a free feature now paid, I addressed in the same release some common user requests like deleting comics from within the app or dimming the screen for night reading.
Late December 2012, ComicFlow 1.1 was released with the new but limited-unless-purchased web server along with other enhancements. I tried different IAP prices from USD $2.99 to $4.99, and rapidly decided to settle on the latter.
The Year 2014 in Numbers
Some high-level metrics to start with:
- 153,471 new app downloads (along with 702,216 app updates)
- 185,000 New Users (not sure why this number reported by Flurry Analytics is 20% higher than the official App Store downloads — maybe due to app reinstalls?)
- 7,900 Daily Active Users on average
- 23,200 Weekly Active Users on average
- 11.5 million Sessions (i.e. one use of the app by a user)
- 15 million comic files opened
- 352 million comic pages read (yes, that’s 3 — 5 — 2)
- 2.3 million comic files uploaded using the built-in web server
- 148,000 prompts to rate the app followed by 11,000 decisions to do so (the vast vast majority of users tapping “Later”)
And finally, the financials:
- 4,440 IAPs sold at the USD $4.99 price-tier (excluding 8 refunds)
- USD $14,623 in net revenues (after Apple’s 30% cut)
A Quick Look at the Past
Since ComicFlow has already been around for 4 years, here’s some additional information to put the above data in perspective:
Closing Thoughts
Some astute readers will certainly conclude that, with a 2.9% conversion rate from new downloads to IAPs sold, ComicFlow is not monetizing very well — especially if you consider how active its user base is.
The truth is that ComicFlow was never designed to make money. It kind of stumbled upon it.
The app is fundamentally usable for free (free as in truly free, not “free with ads”) and most users choose to take advantage of this. The only difference the Web Server IAP brings to the table is a bit more convenience to import and organize comics. Otherwise, one can use iTunes, or even any iPad app capable of “exporting” files to other apps. A number of ComicFlow users for instance use Dropbox to transfer their comic files.
I’ve also never done any optimizations to increase revenue like experimenting with which features should be paid versus free, fine-tuning the App Store keywords, running Google ads to promote the app, translating it to better serve non-English markets, etc…
On top of that, ComicFlow has been entirely open-source since day 1: anyone can download the source code from GitHub and run the full-featured app.
In my mind, the philosophy behind ComicFlow is very straightforward — it offers users this simple deal:
Here’s an app that does one thing and does it well: reading comics. If you want to have a slightly more convenient experience, you can pay $4.99 for it, or you can keep using the app for free.
That’s it. There’s no explicit or even implicit promise that ComicFlow will continuously add new features. If you want the latest comic reading app on the block with fancy animations and crazy organizational capabilities, it’s not ComicFlow. What you see is what you get, but on the other hand, it’s intended to be a reliable and rock stable app.
As you can see in the version history, ComicFlow has essentially been in “active maintenance” for 4 years with a small release every few months at most. Note that maintenance work can still take quite a bit of time and effort like when updating the app for Retina displays or to the latest iOS 7 user interface design language.
Excluding the initial development cost of ComicFlow, which was never intended to be paid for anyway, once in its current active maintenance mode, the revenue per worked hour is quite fine as a low-involvement side-project. Especially if you consider that, as ComicFlow’s first user myself and a daily user of the app, the incentives are here for me to ensure the app continues to work great.
Data Sources
Crashlytics, Flurry Analytics (using custom events where applicable), AppAnnie and appfigures.
