I thought it might be useful to have a way to compare statistics between networks. I know these types of things fluctuate and are affected by a number of factors, but I think there may be some value in at least a broad picture. I put together a site to allow people to enter in their data which is then averaged on the front page and broken down by network and type (banner/interstitial/video). Right now it doesn’t account for geographic region, but I may try to add that if I can come up with a simple way to do it. It does require registration to enter data, but it is very basic requirements and is needed to help people enter data quickly and between different months. I would like to add more features in the future such as more tables and charts, but at this point I am interested in seeing if other people would find this useful and are willing to share their data. The url is www.mobiledevnetwork.com
I am very interested in feedback. I haven’t shown this anywhere else yet. Like I said, I am pretty new to this type of thing and I am hoping the information this provides will help both me(an aspiring mobile dev) and you. Please let me know if I am making any bad assumptions with the way it should work or if you think it won’t provide accurate information for even a broad picture.
While there definitely are variable factors to be considered (app/game type, level of engagement, type of user etc), this is a great idea and could be a very interesting source of information.
I am in.
Thanks, guys. Do you think there is anything I should change before posting this on other forums? Does anyone think its a bad idea? I have some other ideas to go with it, but want to make sure enough people are interested before investing a lot of time into additional functionality.
Its probably best to remove the “app name” stat as most people don’t want anyone to know the name of their apps, especially if income data is publicly available for it. I know you said that’s hidden, but people can’t just take your word for it.
Instead maybe just have a stat like ‘is it an app or game’ and ‘what category’ and make it public. I think something like that would be much more welcoming and it still gives a good idea as to the type of content the ads are being shown in.
One quite large issue is that there is no way of approving developers before they post data. This could be heavily abused by other networks i.e They could sign up and post invalid data that makes their eCPMs and fill rates looks better than they are.
I can’t see an easy way around that issue…
Do it on invite basis and approve each submitter by looking at the proofs provided by him (reputed forum members may be). The quality of data can easily get abused.
TLDR; If these things mentioned prevent the site from being used, the answer is automation. See my response to Toxic for more info.
I thought that might be an issue. I don’t actually care what the app name is, the reason it is there is that I expect most developers will have multiple apps. My hope was that they would enter data each month. It generates the fields automatically for you based on that data and it is only a way I can think of for the user to know what it is referring to. Without having an identifier for the app that the user recognizes, the generated fields won’t make sense. I suppose I could remove all the generating functionality, but then past data would be difficult to edit, I can’t autofill, and it would be a little more difficult to track where you are if you are entering many datapoints. The up-side is that I could remove the registration fall-off. I am open to suggestions here.
Yes, you are right. I was planning on looking into doing some type of verification in the future if enough people are using it - I would most likely need an app name at that point though. Just starting out I am not too worried with it. For now, I was just going to manually monitor outliers/multiple UserIds with the same Ips/User’s with an unreasonable number of apps/etc.
Not a bad idea. I was hoping to get a little momentum before verification though.
I hope to add more charts/tables in the future if it has any uptake. I do currently have a field for entering impressions, but I am not doing anything with it currently(other than saving it).
That is what I feared. It is difficult to enter data. It is work. I have considered automation and it would be much preferable to the current solution as it would solve all the above issues and more, but it is not without problems. Firstly-I don’t want such important credentials stored on my server - even encrypted. I would probably make an app and encrypt the credentials locally on a users device. It could then send just the important data without any identifying information. Secondly, I don’t think all ad networks have API access. I was reading the mediation FAQ on admob the other day and found that they use other networks APIs as well as screen scraping. Any new networks will have to have some sort of adapter written to get the information. Currently, a user can just enter a new network by typing the name. On the plus side, it also solves the problem as relating to geographic regions. It would have more data, it could have daily statistcs in addition to monthly, and it would be easier to use. Programming this will likely take a lot of time, I’m going to have to learn many APIs and do screen scraping. This is actually what I was referring to yesterday when I said I “want to make sure enough people are interested before investing a lot of time into additional functionality.” It would make it much more attractive I think, but I wanted to see the reponse to the existing site first.
Let me ask this question-
What networks are important to you - there are hundreds. I suppose I could do a few of the big players first and gradually add others as requested. I could possibly write an architecture that will allow others to submit an adapter for their network(which would then be reviewed).
I think it’s a good idéa, I submitted some stats. Found a small bug: After already submitted some data and hitting the link “Submit Your Stats” again, all the impression are gone.
+1 for the invite-only basis to maintain the integrity of the dataset.
@mobileguy along with geographic regions, maybe you can add the month that the dataset was collected so as to track the eCPM fluctuations of a given ad network over time.
I think it could be a nice open-source project so that we can contribute as well
Thanks for the input guys. @mak you can select the month on the header of the table.
I am moving forward with automation. I am thinking that the current system is too much work for people and I will start automating Admob and iAd. It gives much more data to chart anyway. I will try to get an APK out with those two networks on Saturday if anyone wants to test it. I probably won’t be available Tomorrow.
The idea is good. But what is the guarantee that you are not yet another ad network trying to build their strategy around what people share around here?
There have been quite a few attempts on these lines before.
I imagine an ad-network would not want a site to exist which tries to compare things like eCPM. It creates competition and competition benefits consumers, not businesses affected by the competition. Even if I was, which I am not, wouldn’t it still be beneficial to you?
I created an app not long ago and needed to figure out which ad-network to use. There are so many, but it is difficult to compare without metrics. I tend to over-analyze things and I want as much data as possible. I can find metrics shared by people in blogs or forums, but the data is always old. Right now, one of the latest threads in this forum is someone asking if others noticed admob dropping. If this idea works, people don’t have to ask. They can just look at a chart. That is why I am working on this. It is something I needed and I figured it is something others needed as well. I am going to keep building mobile apps until I can do it full-time and I do know I will continue to need this type of information.
