The Internet has mountains of Android source code, but not so much Android app financial information. Sure, there is information from the training spruikers telling our kids that if they learn to code they will get rich, because “Crossy road made $10 million in 3 months”.
Everyone not making money seems to quote the number of downloads, like this is in some way a measure of success.
Anyway, it is about 1 year ago that I started developing my main Android app, “Phone health check – and Test”. Rather than sharing source code, I thought I’d share my statistics and financials. It is but one story – I hope it is of interest to you.
Who am I?
I started an Independent Software business, inPocket Software, in Australia in August 2015. Prior to this, I developed Passmark Software’s BurnInTest for Windows and other Windows-based test tools for 10 years. I also had ‘played’ with Android development for a couple of years. Earlier I worked for Mobile operators in the product technology space.
With my background and the huge number of Android devices that will need repair and resale, it seemed sensible that my first ‘big’ app on Play store should be testing Android devices. So, that’s what my app (andTest) does.
I started development of andTest on 30 November 2015 (but had been thinking about it for years). In that time:
- Versions: 32 versions from 4 February 2016 to 22 November 2016.
- Total downloads: Just over 50,000 (fairly consistent daily numbers).
- Current installs on active devices: about 6,000
- Ratings: 196, average 4.2
- Email responses to users: about 300
- Time devoted to andTest: about 6 months (FTE). (I also developed a few small Windows apps over the year).
Huh, some apps seem to have incredible numbers?
So over the last year, I have received numerous emails with schemes to buy or trade downloads and 5 star ratings. I have not used any of these schemes, thinking the best way to succeed is to develop a good app. I do note however that my statistics are not in line with some apps in the same category. For example, I have about 1 rating for every 250 downloads. Some apps in my category, Free Tools, seem to have 1 rating for every 25 downloads. Yes, Google also emails me and tell me that the above practices are forbidden.
It does seem that maybe I have done something wrong here. It seems to me that if you don’t have a rating above 4, then people won’t download your app. So this is really important.
How do people think I make money?
I like asking everyone how they think developers make money from their app. For the most part, they don’t think about it. When pressed the most common answers I get are:
- Google pays the developer to include the app in the store.
- Google pays the developer per download (like $1).
- Displaying ads.
How do I make money?
Sometime after I started inPocket Software, I read an article by an Australian financial journalist, Alan Kohler, about the changes to advertising revenue over the last few years, and that advertising rates (for online news) had reduced by about 25 times. I think it took me a number of months to believe the significance of this.
I basically get paid by:
- Users clicking on ads and interacting with the advertiser’s content;
- in App purchases (to remove ads and add additional features).
How does advertising work?
Users I talk to assume I get paid something like a flat rate for ads that I choose to include in my app.
In short, I use Google Ads (AdMob) and advertisers bid (like an auction) to advertise in my ad spaces in my app (based on information provided to them by Google). While I can remove categories of ads, this means there will be fewer bidders and my outcome may be less. So, no, I don’t choose my ads, the advertisers choose the ads based on what they are prepared to pay. (I do restrict some ad types). Ad rates are also country specific, as advertisers are prepared to bid more for ad space in the USA (for example), than in a country with a much lower national GDP.
This auctioning of ad space sounds like a good thing, getting the highest bidder. I would suggest reading this financial media article I mentioned earlier.
How much money?
Users clicking on ads: andTest has served up about 215,000 ads. Which sounds like a big number, but it is tiny. Really tiny. Of this, there have been 1464 click-throughs on the ads. About 1 in 150 ads are clicked on, or about 0.69%. The average cost per click is A$0.17. Or about A$250 (Australian dollars, 1US$ = 1.35A$).
Don’t I get paid for just displaying an ad? Well, yes, so far about A$5.
In-app purchases: I added in-app purchases about 3 months ago. So far, out of the roughly 6000 installs on active devices, 28 users have purchased the in-app purchase, which is currently about US$2.50. My take from this is A$43.90, or about $1.60 per in-app purchase.
What type of ads?
The general information I have read on the Internet is that interstitial (full screen) ads have a much higher click-through rate than the large banner ads I started with (and still predominantly use). So I added full-screen ads. So far, the click through rate and Cost Per Click is not that different to large banner ads.
I do note that full-screen ads seem very popular with developers (but not users). Again, maybe I am doing something wrong here.
Offers will come along
Some people ask me if I get offers from companies. In my previous life, we got approached by companies asking us to modify and rebrand our software, so it seems reasonable this would occur.
And the answer is yes, I have had a number of offers from small to medium sized businesses. Small businesses have asked that I extend my app for their business, this has generally been a couple of weeks to a couple of months work, for “dozens” of dollars.
An Android manufacturer asked I rebrand andTest and include it on 100,000 devices per year. Sounds promising. But here is the rub, it is a free app, so that’s what people expect to pay. So to spend a couple of weeks rebranding andTest for a manufacturer and assisting with firmware integration, I would get to leave my ads in the app but receive no payment. I made some assumptions about the percentage of users that would use the rebranded app and applied the advertising numbers above.
I didn’t think it made sense for me to pursue any of these offers, so I didn’t.
You developed the wrong app!
After telling people how much money I have made, I get the response “well you have developed the wrong app. You should have developed Pokemon Go or a health app”. Well, yes, I should have developed Pokemon Go … but unfortunately, I am a mere individual mortal. As for health apps, my favorite non-Google app is Strava (cycling), I read a bit about their business and was shocked to learn that they employ almost 100 people .. wow, I would never have guessed that. Lots of people I know (including myself) use Strava, I don’t know anyone that pays for it. But if they have deep pockets, keep it coming I guess.
But yes, my biggest mistake was my app choice. With most people now conditioned to expect apps to be free, and ad rates at the levels described above, I think an Android app needs to have the following characteristics: Daily use and be addictive. My app is neither.
What are you doing, you need to charge for your app?
I have been lectured by friends that I need to charge for my app. Well, just about every app in my category is free with ads and in-app purchases. It is hard to see how large numbers of people will use a paid app in this environment. But in any case, I have includued new functionality in andTest as a purchase. The in-app purchase numbers above would seem to suggest that people are not prepared to pay for an app. Further, I do plan to release a new app in the next couple of weeks – I will make it a paid app ($0.99) and look at the numbers over time.
Summary: Year 1 of andTest
Gross revenue: A$300 (US$225).
Costs: A$4000 (test devices, business costs) (US$3000)
Net position for 6 months work: A$3700 (loss).
Year one has been a bit “it is what it is”. I have enjoyed the experience and would do it again if I had my time over. I have learned heaps.
I have a new app on the horizon. Then a holiday. I will continue with my current apps but look at also including some contract development.