Ad networks and stores pumping up the numbers?

Hi,

So I would like to present the following stats and would like to know if anyone else noticed this.

Case1:
Few months ago I published one of my games to the opera app store. I have a tracking system in place that records each level my users have completed. So far there are 4 levels completed, all coming from the same device. So there has been only one user playing my game according to my stats. However, according to the opera dashboard, there are 73 installs for my game. So why do I only see 1 user playing my game out of the 73 that installed the game from the opera store?

Some might say they downloaded the apk from the opera store and installed it on a device without active Internet connection. That is theoretically possible, but 1 out of 73?
Some might say they installed the game but never completed a level. Also possible theoretically but how many of you installed games without playing at least one level to see how it works? Isn’t that why you installed the game in the first place, to see how it is? I should add that completing a level is very easy and straight forward, you don’t need to be a genius to do it.

Case2:
Recently I published a game on the slideme network. It has 602 downloads in 2 days. My tracking system shows only 100 unique users. What’s up with that? Slideme has an android marketplace similar to google play where you can download and install the game directly on your device. I believe that is where most of the installs come from, so the user does have an active Internet connection.

Case3:
Few months ago I was running a campaign with appbrain to drive installs. The dashboard on appbrain refreshed every few minutes so I could actually the installs in real time. Still, my tracking system was showing only 5-10 unique users but appbrain was showing 30-40 installs. My tracking system is updating instantly: every time a level is completed a request is sent to the server to write to the database.

So I don’t know about you guys but it seems to me like these networks are pumping the numbers up to increase their popularity. The most disturbing thing is that I noticed this with appbrain where I was actually paying for all the installs they said I was getting.

Anyone noticed anything similar?

SlidMe for certain has some downloads generated by crawlers.

Thanks for sharing your research. I never got the trust to waste my time with non-google stores except amazon. I dumped amazon long time back and I see amazon struggling to bring developers to its app store.

Update for case2: 31 more downloads on slideme since I posted this thread but my tracking system doesn’t show any levels completed.

Another update: 7 new downloads and no levels completed.

You have to remember that some people will download an app, look at the first screen, and without even clicking a button uninstall it straight away. Does your tracking system take into account that kind of thing?

I am guessing most of the users downloading a game will try at least it’s first level as long as that level is easy to get to and complete. Why would users give up after looking at the first screen? The user interface, including home screen is included in the screenshots of the game which are usually decisive when it comes to installing a game.

My tracking system only records completed levels, not openings of the game. However, I can check one ad network I am using that tracks the unique users per day and I can say the stats are very similar to my own tracking. For example, there were about 100 unique users opening my game yesterday and about 300 installs from slideme play store.

Well, there are some strange things going on for certain - for example Google Play shows for one of my new games 1000 installs while AppBrain lists only about 50. AppBrain should show even more downloads than your method because it doesn’t need any levels to be solved to count a download. Maybe so many users are turning 3G/WiFi off before starting the app?

I’ve noticed the appbrain listings take a while to update. I’ve noticed 1-2 days delay on the reviews so I am guessing the number of downloads will take same time. So I am talking about the actual listings, not what you see if you’re running an advertising campaign (these are almost instantly updated).

Theoretically it is possible that users turn of their internet connection before starting the game. But so many of them? That is hard to believe.

Maybe you are right. The download stats might also be behind on AppBrain… (although they dropped after a mistake I made in description just a day after, maybe it was not related though and I overestimated the effect of that mistake)

Frankly, almost everyday my flurry stats show less installs than google play. So maybe users are disabling internet before playing cause they don’t want to see ads?

Ok, but how many installs does flurry show compared to the installs showed by google?

My stats show that only 1 of 6 users have an active internet connection if I am to go by that assumption.

About 80 - 90%

Well, that makes sense! But I am seeing 17% in my game.

This Opera store almost looks like a scam, so I wouldn’t be surprised at all that they did inflate the download numbers. They keep sending spam to my games’ support emails on Google Play (that means they are scraping Google Play for emails) asking to join it.

On the other stores (Amazon and Samsung), the numbers look legit to me, they seem to match Flurry’s numbers (or are pretty close). When there’s a peak on one of the stores for some reason, on Flurry it peaks respectively, so I’d call them trustworthy.

So I just released a new game on Slideme market and they say my game was downloaded 74 times (in about 2 hours or so). However, I only have evidence of 2 users that actually played my game.

I have experimented the same situation for my first game with Slideme, lots of downloads but only a small % tracked by flurry or generating revenue. I thought about users turning off wifi, but as you mention , it should be a small %, not the majority of users.
So i forgot about Slideme and other secondary stores and never looked back. Not regretting at all, specially now with the launch of Google Game Services and other GP-exclusive features

Actually I do have GGS in my game and users are presented a popup at start asking if they want to sign in with google. So looks like GGS is not forbidden on Slideme even though they specify most of their users don’t have access to Google Play market.

I agree with you that this market is not worth it. But I like releasing it there first before I release it on google play, it can help me track problems/bugs before I jump to the important market.

SlideMe allows direct APK download from their website, so probably spiders download a lot of apps. On the other hand… good apps seem to get much more downloads than bad ones in the first days after publishing there, so I don’t know. maybe SlideMe artificialy boosts the downloads or something. Anyway I had a success in alternative stores only once - when I got a lot of downloads on Amazon during Christmas for some reason and only on one app.

If I am not mistaken, mind had a boost of downloads around Christmas too from Amazon. His explanation was that there were a lot of new Kindle Fire devices purchased around that holiday and there weren’t a lot of applications in the market.