“In the very beginning, I had this foolish idea in my head that this flood of people would automatically rush to my website, buy my product, and I’d be a millionaire within months. Reality struck when I had a whopping 5 visitors to my site in the first month. At the time, I didn’t understand that you actually have to DRIVE traffic to your site, as people won’t just magically find you.” – Mike Geary, founder of “The Truth about Six-Pack Abs” from his interview with Tim Ferriss
I had what they call a “holy sh*t” moment the other day.
If you’ve been following Bluecloud, you know that I launched Coconut Craze last week. I’ll be writing up the full breakdown of what happened with that shortly, but it’s way too big to begin yet.
The launch of that app had many discoveries – the truth about paid vs free apps, why Lite -> Paid is a dying framework, and why advertising has quickly become the biggest opportunity I’ve ever laid my eyes on.
I’m talking hundreds of millions of dollars in profit. Eight zeroes.The kind of money that puts you into a league of people that has since been reserved for a few.
I’m here to show you how you can make that sort of money.
App Marketing 256 – Revenue Modeling
Before I get into it, I need to define a few things so that we’re all on the same page.
CPC (Cost Per Click) – This is one of those great terms that’s been floating around since Google AdWords and other web marketing advertising channels gained popularity. It’s exactly what it sounds like – you pay every time someone clicks on your ad. You’re buying traffic, as each click translates to one visitor.
CPI (Cost Per Install) – This is a term that is fairly recent since it deals directly with apps and is the modern version of the CPA (cost per acquisition) often used in lead generation type marketing. CPI means that you only pay when someone downloads and installs your app.
Chartboost – The advertising platform that I began using with Coconut Craze and have installed on most of my other apps. Instead of being the traditional banners like iAds or Admob, their preferred platform is a popup that appears when a user boots up the device (or every 24 hours, etc). They offer a full analytics platform to track everything you can imagine. It’s awesome.
eCPM – Probably the most important metric in app revenue modeling, along with all advertising in general. eCPM stands for Earnings Per Thousand (impressions). For every thousand people that see an ad, you can expect this dollar amount. It shows how effective your platforms are.
Revenue/Download – This is just a simple metric that I use to calculate what your average download will bring you over the lifetime. When you aggregate all your revenue streams (in-app, ads, playhaven, chartboost, etc) and divide it by the number of downloads you received that day, you will have this number. For example, if you have 5,000 downloads and made $500, you would have a R/D of $0.10. For every person who downloads your app, you can expect to make, on average, $0.10 from them.
Yumm. Don’t those acronyms make your mouth water?? Or is that just me?
On to the good stuff.
Buy Me a Sliver (of your traffic)
When I first opened up my apps to Chartboost, I wanted to test everything as fast as I could. I’m the kind of person who learns from experience, so the more I could see, the faster I would be able to see the trends and create conclusions that would lead to a strategy.
The apps went out and started performing well, at least from my perspective. Alpha Combat, for example, was pulling in a 35% CTR and a 18-20% Install Rate, leaving me with an eCPM of about $75. I chalked that up to being a new update and having all the users interface with the ads for the first time.
I started thinking about being proactive though. With metrics like this, why don’t I just buy downloads to run through this funnel? I do this by simply reversing the traffic flow – instead of advertising other people’s apps, they advertise mine.
I dropped $200 into my account and set up a quick campaign.
There Are Rules, Jerry!
A few hours later I receive an email from Chartboost’s customer service (which TOTALLY kicks ass, btw!), letting me know that I can’t actually do a CPI campaign on a paid app. The reason is because it’s not fair to the market if they’re bidding against free apps (at least not in the framework that CB has up right now). Thus, I have to do a CPC campaign for paid or a market a free app for CPI.
So I switched it over to CPI and locked down Alpha Combat.
My minimum bid was $0.50, which is the minimum for the network as a whole. For every installation that publishers in the network gave me, I would shell out $0.50.
I watched this unfold at lightning speed, as my $200 spend is nothing more than a blink in the world of digital media buying. In a matter of minutes I had 10 purchased installations.
Awesome, right!? That means in the next few hours, hopefully, I’ll see a similar uptick in my revenue for Alpha Combat.
Such an amateur move.
It took me a night’s sleep to realize how blind I had been to something so obvious – I never calculated my Revenue/Download metric on the app I was advertising. Yes, the metrics were awesome. Yes, the app makes a lot of money.
No, I had no idea how many people WEREN’T making me money (people who downloaded the app, that is).
Calculations, Permutations, Summations, and Sexy Equations
I paused my campaign and pulled up a month’s worth of historical revenue data for Alpha Combat. I kept things pretty straightforward – add all the revenue (I did this by day, you can do it by week or month if you have enough data) and then divide by the number of downloads for that same time period.
I thought I was so smart. I couldn’t wait until that beautiful $1.02 metric appeared.
Yeah right. More like $0.08.
Every time that’s what it came back with – $0.08. For every person who downloads Alpha Combat, I can expect to make $0.08.
More importantly, this means that for every CPI install I paid for, I was basically giving away $0.42.
I’m almost embarassed to admit how easily I lost myself in the grandeur of opening up an advertising funnel on one of my apps, but it was quite the lesson learned. My apps have no business being advertised.
….unless they make $0.51 Rev/Download.
Hitting the Gas on Penny Lane
I stepped back and realized something that is SO obvious but I wasn’t able to see it clearly – I can’t advertise in a CPI network until I have an app that makes more than the CPI cost.
CPC is a much harder metric to forecast and measure because the conversion rate (install) is still variable. Too many variables in a framework means trouble in business. I’m sooo over variables 🙂
Looking at the “Apps That Are Advertising In The Network” list, I started to see a trend – these are all very complex games that have some sort of awesome currency system built in. Even if it’s not a literal currency, almost all of them give users the ability to spend a TON of money in the game.
Similarly, when you look at the Top Grossing apps in the app store, you’ll notice that most of them have the same characteristics (the free apps, that is). They’re also the games that I see most often in every ad network across the entire market.
The defining difference between apps that CAN and apps that CANNOT pay for installs is whether or not their Revenue/Download exceeds the CPI for that network. Once you achieve that level of user monetization, it’s party time. You can go get your Amex Black Card and spend $50,000/day and make $51,000. If you’re beating the $0.51 benchmark, your upside continues to rise. Some of these apps (from what I’ve been told) have Rev/User metrics in the order of $4. Spend $50k, make $200K.
That’s when an app’s success becomes less of a marketing exercise and more of a financial exercise, and when you turn a business model into a financial model….well, that’s when you become REALLY rich.
What Does This Mean? What Should You Do?
What this experiment taught me is that I’m going to be owned by Apple and old school hustling/marketing until I can control my revenue stream.
Yes – there are some awesome ways to increase your revenue/download before you need to break that $0.51 barrier – push notifications, updates, multi-player incentives, price manipulations – but these are still dependent on HAVING a lot of users in the first place.
The takeaway you should have from this article is this: start looking at your apps like investments. Approach every decision the way you would if you were running a company and back it up with data. Measure EVERYTHING and optimize every hole you have.
You’ll start to see pieces of the puzzle that you never thought you would see.
Once you take your app and really break it down to metrics, you can begin testing the app to see what changes and what the effects are.
“An update can expect to re-engage 40% of the users, who will purchase 22% more than usual, leading to an increase of $XXX in revenue for that particular update. It’s going to cost us $XXXX to build that new functionality. Yes we should do this.”
My own personal strategy is to build an army of games that all make anywhere from $5-100 a day over the next 6 months. I already have about 5 of these and am averaging $175/day. This needs to be 15-20 games averaging $400-500/day. THEN I’m going to sit down, hire an awesome developer for 6 months, and build an app that’s going to be as close to the $0.51 rev/download metric as possible.
I hope you find this helpful and provides some new insight into how to become a major player in the app world.
Talk to you soon,
Thanks Carter that was helpful
Good information. It makes total sense. We’re working on a similar strategy and having a blast.
really inspiring story
Very good post….interesting that I planned out my 12 mos on the plane back from the AE event last weekend, and my strategy is similar – and just now I am reading this post! Awesome! I will refer back to this often, Carter.
Right on, glad you liked it. It was great meeting you the other weekend as well. You’re destined for great things – you have incredibly positive energy in you. You’ll hit your stride and the rest will be history 🙂
Awesome Post, Carter! This gives me a bit more clear path in planning the next 6-12 months in building my own App Empire! Awesome meeting you at AE a few weeks ago, btw! 😀
Thanks man. Great meeting you as well. Keep in touch!
Great article. I appreciate the honesty in the numbers, and I’m glad I’m on a very similar strategy… It’s working awesome for you. And I can’t wait to see that big gun you’ll come out with!
Thank you for helping me. I just recently found all your posts, and as an amateur app maker I am wanting to develop a few ideas and hopefully have them pay for themselves. I am working hard and taking in all the advice you give. I read your site almost every day and I am hoping to learn from you and someday be just as awesome as you.
Kick ass man! Keep rocking. You’ll make it happen.
Unable to understand a few things:
1. Did you make the app free to leverage the CPI campaign ? If you did then will this not affect your revenue/download, the metric you are tracking ?
2. With the $200 that you shelved out, did you get all the 400 downloads ($.50 / download) through CB ?
1. It changes it. I made the game free, then tracked organic vs paid traffic for the time it was free. Not an exact science, but enough to know what’s working.
Thanks for reading!
Just thought of something. With CPI advertising, unlike CPC, it’s not necessarily to your advantage to maximize install rate. Instead, it’s to your advantage to discourage installs from users who are unlikely to spend money on your app, to prevent them from bringing down your revenue per install. Have you thought about identifying the characteristics of your paying users vs. nonpaying ones, and designing an app page that appeals to the paying ones while being deliberately off-putting to the nonpaying ones?
I’ve saved this article and refer back to it… it’s great, and makes perfect sense. If your revenue/download rate is higher than $0.50, you should take a loan for billions of dollars and spend it all on advertising. This has definitely changed my thinking.
Unfortunately I read this after I had spent a huge portion of my savings developing an app that doesn’t monetize. I also don’t have the skills to develop my own apps or to re-skin. But what I do have is a marketing machine, and my current app has stayed in the Top 250 of the App Store in its category since launch with my own time and no marketing money. How? Marketing through massive social networks on accounts that have gigantic followings (100,000+ followers).
So if anyone’s out there that is interested in making money and has the other piece of the puzzle–an app that can make >$0.25 per install, find me. We’re in for a ride.
@Justin – have you tried re-skinning source codes? I’d recommend signing up for that list I have going. It’s a great way to make money and keep costs low.
I’ll check it out. Thanks for the quick response!
Great article, thank you.
(3 months later)
What’s up Carter! In the couple comments above, you recommended I look into reskinning apps. My partner and I took your advice and started powering through your source codes and have 9 skins completed (many waiting review). It’s a really fun process and I think it’s awesome that you’ve shared all this info. We’re approaching 15,000 downloads on your reskins off only 3 games, so it’s exciting, and best part is that we’ve been able to cut costs by doing them all by ourselves.
An idea popped into my head–it would be great to have accompanying video tutorials to some source codes. I might be able to put some together for the codes you offer and break it up (implementing ads, doing in-app purchases, etc.). I was wondering if you think that’s a good idea and if you’d allow it for certain codes.
On your side, I think it would help the people who have no technical experience (my partner and I have very little and still managed to figure it out) and they’ll be more likely to try their hand at it without having to hire a programmer and/or graphic designer. For my company, it could help our blog get exposure with great content and establish us in the app reskinning world–especially if you’re linking to us.
Feel free to message us at pixtant at pixtant dot com. That email might look familiar as an Appflipping Member on appempiresoftware & flipsourcecode!
Thanks man, and we look forward to more of your source codes.
Funny carter, your most recent e-mail took me to this post. It’s actually something that logically just came up in my head and I started playing with numbers, and later I went on to do ‘Google’ searches on improving revenue per game and downloading all the top grossing games to see what they all do.
If you want links to some of these websites of people who discuss different paid traffic methods, you can see my e-mail, just send me one and I’ll send a list of a couple of places I’ve been reading on. Perhaps you have them already, perhaps not. But they helped.
But after confirming my thoughts, I started working out that almost all source codes have their ARPU, no matter what theme is used. Some themes get more downloads, but in the end all source codes that I released provided an average that were similar to each other.
For example, the ‘Crazy Poppers’ code produced roughly $0.07 ARPU from worldwide traffic, whereas when you focus on just US/Japanese traffic it went higher at like $0.15 per user. This was without modifying the Ads that were being shown to users so those stats could have been improved.
Using this data I knew that the way forward with reskins is to use them as ‘Starter’ templates, and that I would have to work on them and improve them to increase the ARPU. New levels, functionality, and more engagement methods.
That wasn’t the only useful thing I learned, but I then learned that by using this data, I could work out at what price I would need to be able to reskin specific source codes for profit, how long it would take to make the money back and how many downloads I would need to achieve daily to make the money back in the time I wanted to. All this data really helped and I’m able to work everything out in advanced pretty accurately and the difference in results is usually very minor.
It feels a lot more secure and safer when you can begin working based on the data you have collated, whereas when you first start you don’t have this data, which means you have to invest/risk what you have to obtain it.
Ideally it sounds like you’re heading in the same direction. Working on several source codes that you can buy traffic for at a profit, and then going crazy with ad spend 🙂
This isn’t to say that this is easy, a lot of A/B testing has to be done and apps may need to go through several updates before profit can be made (funnel needs a lot of tweaking). But once you have, you can replicate that success with that source code many times over for different themes.
It’s nice to know that industry leaders are on the same page as you, it just adds to the confidence that I had in this specific business model.