Income Report - High CTR, Low $

Hey all - this forum was incredibly helpful when figuring out how to monetize my app, so I thought I’d share my experience as well.

My app (Song Key Finder) was released 3 weeks ago. My goal is to get it to a point where I can make back my hours in income, though I’m piddling along at the moment.

The app works by recording a sound sample with the device’s mic and returning the key that sample was in. The audio analysis occurs after it’s recorded, so there’s essentially dead time for the user. I used this time to play a video ad (rather than go with an app pre-roll) because it’s less invasive and had potential for multiple impressions per app use. I also have a banner ad at the bottom.

Distribution after 30 days
SlideME: 990 dls (free)
Lots of initial dls (seem to be a lot higher for games, almost zero for paid apps). Not sure if updates are pushed to these users or not. Also, users may not be in countries where networks like AdColony are serving ads.

Google Play: 320 dls (free), 3 dls (paid)
Analytics are good, and potential for Play Store cross-promotion is good. Getting 40-50% active install rate.

Analytics
Playtomic
I used Playtomic for analytics because it was free, but it had a few minor issues with location reporting, and it took me a while to dial in how the analytics work. It has been a useful troubleshooting tool, though, to track ad network behavior and fill rates. I’m getting a 1.5-min average app session length.

Video Ads
I’m getting about 0.5 video ad impressions per app open. So, less impressions that a pre-roll, but I hope the increased engagement will add to the lifetime value of a user.

AdColony
I love this network ($5-10 CPM, pay-by-impression, and fantastic support), but the videos are getting played only 5-10% of the time. I’m not sure if this is a fill rate issue or if my SlideME downloads are in territories their advertisers aren’t buying. It could also be a bug on my end, which I’m dreading, but I’ll keep feeling it out.

Vungle
Vungle is my fallback, and it’s played videos almost every time AdColony can’t. The first couple days saw a $5-10 CPM, but it has since dropped to $0, pulling the average down to less than $1. It’s CPI, so my only guess is that my users aren’t that interested in installing games, or possibly they’d rather get their key finder results than leave the app and install another one.

Banner Ads
Revmob
I use Revmob interstitials as a video ad fallback, but they’re rarely called. I use the adlink ad type in a row with “Rate” and “Upgrade” buttons just above the banner ads. Currently this button gets a 5-10% CTR (per app open, essentially a 90-sec refresh). Before I added Mobclix, this button occupied the entire banner ad slot and got 25% CTR. Despite this high CTR, I only have one Revmob install so far, with about 30% of the CPI as Vungle installs and a current CPM of $0.67 (though that’s counting adlink clicks as a single impression with a 100% CTR, so actual CPM is much lower, more around $0.03-0.06). This is a far cry from what I’ve heard from other developers, so it may be that my audience just doesn’t care about installing free games.

Mobclix
I use a banner ad at the bottom of the app. I started with a 15-sec refresh, but after seeing low CPMs (~$0.07), I changed to 30-sec, which doubled my CTR, which is now at an unbelievable 9%. Impressions are still very low (3000/mo at 30-sec refresh), so my CPM is unchanged. Although changing the refresh effectively halved my income, I’m only dealing with pennies here, so I hope the CTR will encourage advertisers to buy space if my impressions rise (I think that’s how it works??).

Conclusion
This has been incredibly hard. 1/3 DAUs click a banner and 1/2 see a video ad. Still, in 30 days, with 1300 current installs:

Adcolony: $0.70 ($7 CPM video)
Mobclix: $0.15 ($0.07 CPM banner)
Vungle: $0.75 (<$1 CPM video)
Revmob: $0.15 ($0.05 CPM button/interstitial)
Free version total: $1.75 (45% active, 3% “+1’d” it)
Paid version total: $10.50 (0.9% adoption vs. free)

So either most people hate the app and click on an ad to leave it (which reduces a user’s lifetime value significantly), or my users don’t want to install the games being advertised on the CPI networks (Revmob, Vungle). I’m pretty sure the low Mobclix CPM is due to my low monthly impressions. I’ll keep you all updated next month!

High CTR might mean that you are always showing the same ad.

How does that work? I probably am. Most of the Mobclix clicks come from a single network, and the Revmob ad unit is a static button. I would think that would lower CTR though? Unless I’m getting very few repeat users and most views are from new downloads.

A bit of an update, 2 weeks later:

Part of the high CTR was due to a bug in some devices that placed the ad over a “Return to Menu” button. Not good at all. I removed the banners, since mobclix was only getting me $0.10 CPM anyway. I replaced that banner space with a larger Revmob adlink. I hope to see the Revmob CTR to increase 8-10x as my Mobclix clicks convert to Revmob clicks (which could result in a $6.90 eCPM for December). We’ll see! Adcolony’s fill rate has also improved over the last couple of weeks.

Usage Analytics
Only 1.25 views per download; hoping solving the banner bug and more time elapsing will increase this. Monetizing at $2 per 1000 app views. That works out to a quarter of a cent per download.

November Income
Adcolony: $1.44 ($6.84 CPM video views, 25% fill) - pays by video view
Revmob: $1.27 ($0.65 CPI / $0.01 CPC / $0.69 CPM button ad link, 100% fill) - pays by install only; numbers are interpolated. CPM based on 90-sec impressions.
Mobclix: $0.40 ($0.10 CPM banner ads, ~100% fill) - pays by impression mostly
Vungle: $0.74 ($1.40 CPM video views, 100% fill) - pays by install
Total Free: $3.85
Total Paid: $14.00

Total Installs
SlideME: 1130
Google Play: 630
Amazon: 30
Appia: 5
Total: 1795

I found a few pieces of info interesting in your reports:

  • one that SlideMe is giving you MORE downloads than Google Play - I’ve been ignoring SlideMe on the reports by folks that the markets generally remain minor compared to Google Play (though exceptions have been reported).

  • second that for 1300 downloads you are getting $10.50 vs. $1.75 from ads - while better than ads that is about a 1% conversion rate (assuming you are charging $1 average in-app purchase).

Couple of caveats to those results: the paid version was selling for 4.99, so there were only 3 downloads (10.50 after Google’s cut). That worked out to 3% of free installs at the time, but these were Play store links in the free app, not actual IAP to unlock. I’d be really interested to see if implementing it that way increased paid sales.

Also, SlideMe’s downloads are HEAVILY weighted toward the first day and week of release. Paid apps seem to get almost 0 downloads on SlideMe. Also, video ad networks like AdColony don’t serve as much to those users, since a lot are outside the US, and I believe networks like revmob won’t work nearly as well, since they’re linked to the Google Play store and most of the SlideMe users don’t have Google Play preinstalled. IAP might work, or networks like Mobclix/Admob.

SlideMe is a great alternative out there for free apps. After 6 months or so, not a single purchased app though. The key I think is comparing competition. On Google you might have 10 competitors, Amazon 7, but on SlideMe 1 or 0. Your app will do better on SlideMe for at least the first couple months. This brings me to another point, SlideMe users seem eager to download something that first month or two and then it dies off after that.

Thanks for the feedback.

Well probably the statistics would make more sense once the downloads reach in 1000s.

How was your experience with video ads ? I know that they reputedly have low fill rate (perhaps doing mediation over a couple of video networks would help).

From what I tested, I found the video players take some time to appear - and whether the video ad is cached ahead of time may vary between the video ad networks …

What’s the payment like for video ads - is it per impression ? And what’s your understanding of the partial viewing/payment situation.

Some of the video ads may allow immediate close (does one get paid for that impression ?) - while some are stuck for half the time (essentially forcing the user - does one get paid for half the viewing time ?).

Perhaps related to how SlideMe presents apps - i.e. new apps being highlighted or something - which would explain that behavior.

Google Play has a very obvious 1 month “concession” behavior - where your app gets taken up in the rankings (perhaps to see where it will “stick”) and then by end of month the app falls in rankings.

If the app is good (or fits some search criteria i.e. matches what people are looking for) it will eventually then start to climb back up in the rankings slightly.

SlideMe has a “new” section. Also when you update your app it’s probably shown it a special section (I always get more downloads after updating). And it’s not only bots downloading like some said - because better games always get more downloads than worse ones. :slight_smile:

SlideME’s website has all apps listed by release date, so new apps have huge visibility. Not sure if mobile interface is the same.

As for video ads, AdColony is pay-per-completed-view ($0.005-$0.01). There’s no close button (you have to quit the app) so the completion rate is around 98%. The fill rate, however, is 10-20%, and CPMs vary slightly based on your CTR (from $5-10; I’m averaging $6). Vungle has a 100% fill rate, but is CPI after viewing the video ad. My Vungle income is about 4:5 vs AdColony’s but it’s also being displayed about 6x as often, and as such has ~$1 CPM.

Revmob (CPI) is now getting $5/week from a static ad link button. I believe it takes a few hundred clicks to dial in the correct apps to link to for your users, which increases install rate. CTR per app view is 80% for me, but install rate of those clicks is only 0.6%. I’ve read it’s usually more like 2% for Android, 6% for iOS.

Tricorder and Idiot Detector are getting 200 and 60 free installs/day now, and about 5 and 0.3 paid installs/day. 20% of free users click on the upgrade button, which goes to Play Store, 12% of those buy paid version (assuming no one buys organically). So, about 1:50 paid:free. Song Key Finder is still 30 free installs/day, with a 1:100 paid:free ratio (though the price is $4 instead of $1).

I’m launching an unbranded Tricorder app (because it’s just a matter of time until someone asks Google to take “Tricorder” off the store) in a couple of weeks that actually uses phone sensors, with IAP to upgrade. No video ads, just a small revmob link and nerfed features for free users. I’m very interested to see how engagement time and free-to-paid conversion changes.

Thanks - that is great stats visibility.

Admob revenue is increasing - perhaps leading up to Christmas. I am seeing eCPM in $0.40-$0.60 range.

Of course the eCPM figure is dependent on the refresh rate (i.e. if you double presentation of banners, that will halve eCPM though you may not be paid any less).

So presenting that in terms of DAU (Daily Active Users) - though here also it will vary by app type and user retention for that app etc.

For about 5000 DAU (Daily Active Users) over the weekend (up from 2500 just prior to the weekend) - as reported by AppBrain stats - I am getting about:

Admob - $9 revenue ($9, $11, $6 over weekend) - and prior to weekend was $3 or so
This translates to 5000 requests per day for admob - peaking to 15,000 or so over this weekend (which was a bit high).

AppBrain - $4.5 over weekend, $2.5 prior to that (half of this is AppBrain banner ads and half is the “More Apps” that is shown at end of app once every 3 days).

Now this is something new - usually AppBrain gives more than Admob - but the higher eCPMs (and jaded response of users to AppBrain More Apps) may have led to the reversal as we approach Christmas.

Generally I have found AppBrain to be a reliable stable source of revenue.

I present a full screen interstitial - still using Greystripe - which is supposedly impression-based (i.e. they pay you for every ad showing). It doesn’t have an (X) (which confuses some users who think the ad is stuck or something) - but if you press at bottom it shows a “Skip” button.

I have been thinking about replacing the Greystripe as it has been paying eCPM of $0.60 (now up to $0.70 as mentioned below) - which is barely more than BANNER ADS - and full screen ads should be much more (!) given the space that is given to them.

Greystripe generally averages $1-2 per day - which is similar or worse than Admob banner ads !! HOWEVER over 2 months there have been two weekends which have made revenue rise to $7 and $9 per day - however this is the exception. But it points to what the revenue POTENTIALLY could be - I suspect Greystripe maybe lacking in ad inventory.

So a good full screen ad network may give much more.

Additional banner ad networks

I am considering adding madvertise and those other networks which supposedly promise higher revenue from european users - but not gotten around to implementing the custom coding - right now just using Admob mediation (serving Admob and AppBrain banner ads as described above - tried Millenial Media but eCPM was same - someone has suggested the default refresh rate on that is 30 seconds and increasing to 60 seconds improved their eCPM - however I don’t know of a way to set the refresh rate when using Millenial Media via Admob mediation).

Alternatives to Greystripe

Now I am looking for alternatives to the Greystripe full screen interstitial - gordon etc. have pointed to Airpush SmartWall as offering a variety of ad types presented depending on internet connection etc. - but I have been wary of Airpush and waiting for their non-EULA SDK - which has been promised but not delivered yet it seems.

So what are the choices for the full screen interstitial (which gets the most criticism from users - for taking over the screen - though I do present it not more than once every 2 minutes and even that is held back if the user is using the screen - so I’ve programmatically designed these to appear only when user leaves screen idle for a bit). In practice the Greystripe is fairly easy to get rid of - i.e. tap and tap on Skip to get rid of it.

I am wary of the video ad - as I suspect they take (from the ones I’ve seen) a long time to set up - it MAY work maybe on startup, but I just found that problematic - compound that with their low fill rates and such caveats and I’ve held back on changing this.

As of now, these full screen ads may provide a function of nuisance perhaps - as incentive for users to get the full app - for that I am trying to get GetJar implemented. However their SDK3.0 is new and they don’t have good docs yet (may later) explaining that - except for a demo app which is slick but is convoluted from developer point of view.

GetJar Gold (GetJar Rewards) and StartApp

I hope to get GetJar implemented - then see what kind of revenue it can deliver and compare it to something like StartApp. I cannot offer Google Checkout from my location, so GetJar is an improvement - but GetJar suggests that the conversion is 4x higher than for Google Checkout (which is abysmal in general as android folks are less likely to pay cash than iphone counterparts).

I am not too keen about StartApp - but I can see how that model can work - some folks here are already using it - and I see apps now on Google Play using that. Oddly very few users complain about it. In any case it is less fishy than Airpush which is already doing notification and home screen shortcuts - at least the StartApp has a very appealing dialog box that essentially pleads user to Accept - and points out the changes that will be made. From showing it to people I have seen that users don’t seem to find it that threatening.

Perhaps users who rarely use their browser would not be bothered by such a change (i.e. add bookmark to browser, change home page and a shortcut on home screen that leads to their search engine).

The only thing that bothers me about it is that I would be wary of such a change - or having something spoil the bookmarks etc. - and would prefer something like GetJar.

However the reality is that the general public may have very different thinking that the programmer making an app.

As it stands I would say that users of my app are probably MORE pissed off at the ads I show (banner, full screen interstitial and AppBrain) than they would be to an app which just showed StartApp at startup (a type of nag screen).

The few users who have reported on it here suggest a 80% conversion rate. Now this makes for a reasonably predicatable revenue model - i.e. 80% of ALL downloads (not just retained downloads). And they supposedly pay $0.05 per SDK install for U.S. and $0.01 for non-U.S. etc.

So in any case I am waiting to see how the GetJar works - and then compare with the predicted calculation of what StartApp could deliver.

StartApp seems ideal for small apps that do one thing and have a simple interface - there the nag screen will eventually make nearly all users accept the dialog box which appears at startup - and so conversion of 80% as reported by some here may not be wrong. If 80% is correct - that gives for some very RELIABLE revenue figures.

So if you have 1000 downloads (and not retained users but the raw initial downloads !) - 80% of that is 800 - each paying $0.05 to 0.01 and maybe $0.018 on average (startapp.com - click on “I am a developer” and it shows their current RPMD at $18 i.e. per 1000 installs of their SDK).

So 800 x $0.018 = $14.40

Now compare that to GetJar - if say 25% of those 1000 downloads are retained users i.e. 250 - and of these say 25% go through and pay the GetJar process - that is:

250 x 0.25 = 62.5 users

And if each paid $1 via GetJar (or more precisely 100 gold coins - which gets you $0.9 as a developer) - then that is:

62.5 x $0.9 = $56.25

So if these percentages are correct you COULD be paid well by GetJar - I will have to see once it’s implemented what the conversion ratios are …

But the advantage of GetJar is that it allows for additional payments.

StartApp has a limit to how much you can charge the user - i.e. beyond installing the StartApp browser bookmarks etc. there is not much more the user can do for StartApp - though in future they may offer revenue sharing option (would be more complicated account keeping etc.).

GetJar vs. StartApp

So I suspect StartApp maybe actually (from a general public being ignorant of things and winding up prefering StartApp - for it’s simple one-click Accept - vs. say GetJar more complicated payment process - earn coins by installing apps - and then pay using those coins) - one COULD split the decisionmaking this way:

  • for simple apps, live wallpapers etc., it maybe better to choose StartApp - as it is a one-step (and user can Decline also) - so seems to give a choice and app is completely open to the user (i.e. “if you wish to pay us” - which may actually work i.e. users may think “what the heck - I should pay them by clicking Accept”).

  • for apps which are more complex (i.e. you want to earn MORE than $0.01 to $0.05 per user) - or where you want to be offering additional features later and charge for them - for that something like GetJar maybe better

I do not know, GetJar do not fill me with trust because their site is very unprofessional and their app have at least 2 pages with 1* rating comments how users do not receive gold. @adforandroidapps eagerly waiting your reports about them.

Yes, as I have outlined earlier they seem in dire need of some good programmers - so while they are spending VC money on getting developers signed up and nearly all apps are earning coins - so GetJar is in the spending phase, they really need to spend some money on quality programmers.

Their GetJar Rewards app is kludgy - though they are releasing updates. And their demo app for SDK3.0 (which seems to be the only “help”) is a bit more complex than it needs to be for a demo for developers.

However, they seem to be moving in the right direction - I think some of the problems also are from their moving away from local preferences to having users’ gold balances being kept at GetJar servers (it is linked to the users’ Google Play id) - so they can get their payment history back if they change phones etc.

For some reason they are being allowed to survive on Google Play - and may even be a possible acquisition candidate for Google I am guessing (after all you will not find a SlideMe app on Google Play - but GetJar Rewards is available on Google Play).

I have yet to get GetJar working in an app, let alone to be in a position to comment on their payment promptness and all that stuff.

However I do not mistrust the company - they are a company who have been in this area for a while - and probably predate all these other app stores - they got kicked out by Apple in part over the use of the App Store name etc. But I have a gut feeling this is a company to watch.

Once I get it running, I can post more info on conversion ratios etc.

The 1 star comments you mention are probably related to transition issues as many of the apps currently using GetJar have not transitioned to SDK3.0 - GetJar is offering an incentive to these developers to get it done by some date this month.

No, I trust them as company after all it is obvious that Google and GetJar have some sort of deal behind scenes, but when talking about trusting I think on possibility that this system really work in practice for now ( base of 20 million users is a bit small in my opinion because when we remove inactive it can drop easily on half ) of course maybe ( and I wish ) that I am wrong. Still waiting for someone to open personal report because I can not find any on the internet.

Well, I am willing to try them - since I don’t have option of using Google Checkout.

And because I have seen AppBrain in action. GetJar cannot possibly be worse than AppBrain - because in AppBrain the choice of downloading a “More Apps” banner ad or interstitial at end of app is totally up to the user.

If THAT can earn reasonable amounts, then certainly a GetJar type (which essentially let’s you know when a user downloads an app - so you can unlock features etc.) should do much better.

So the equivalence between AppBrain and GetJar is that:

  • AppBrain - developer doesn’t know when user downloads an app (they just get paid for it)
  • GetJar - developer knows when user downloads an app (and can unlock content/features)

The difference is that AppBrain is a much slicker company - they seem to run a small tight ship.

Also AppBrain has the advertisers - so most apps they offer are reasonably high quality.

GetJar is trying to jump start it by throwing money at it to get to scale - so right now even if you download Skype you earn 10 coins - same for many other apps - now obviously all these apps are not paying GetJar (yet) - but GetJar is wanting to get to scale probably and fast - so they are willing to pay FOR those app developers (as if skype was advertising via GetJar) - so that at least for the end user - they see that “hey I can download almost anything I usually would and I earn coins as well”.

So I don’t doubt that the model will work - or will work no worse than AppBrain - but what it does is “enable” the developer to do some type of unlocking of content etc.

At the other end you have Tapjoy to compare - it too offers app downloads (though it does other offers also). So users are familiar with that.

The only problem is in execution - whether GetJar Rewards app and other stuff will improve and become responsive apps (they have improved but still are relatively slow).

The other question is if this GetJar model can “scale” i.e. can it expand to fill the ecommerce needs for all apps ?? That is - will the app advertising by the big commercial app houses cover the payment needs of all the smaller app players ?

I had this question before, but it seems GetJar has a solution for this - they are expanding ability (though I have not seen it appear yet in GetJar Rewards) so that users can eventually BUY gold coins with cash also.

Now THAT will essentially open things up - and will remove the “will it scale” issue.