New networks rev-share and performance

Solar2D Monetization (in-app purchases, ads, etc.)
#1

Hi,

Has anyone tried working with one of the new ad networks and can share his/her experience?

I’ve tried contacting AdColony, chartboost and UnityAds to get their commitment for the rev-share they take. As expected I could not get proper answer.

I did find that AdColony terms say they pay only 50% for the first 3M monthly completed views per app! This is very low. unless you have a great app usage, you will not get there. Seems not worth it on one hand but their eCPM is said to be good. Has anyone tried them vs vungle and can say which provides better eCPM, fill rate?

chartboost rep did not want to commit to a certain rev-share so I did not try them yet. their eCPM is said to be lower than AdColony but lower fill rate. anyone experienced with them? (perhaps with the old integration)

Last is unityAds. again, they do not expose their rev-share (at least to me and my brief search over the web). I did read quite a few who claim for cumbersome admin and slow response. anyone tried them?

We have a mix of adult and kids apps on both iOS and android. of course eCPM varies by country and by app type but the overall eCPM is a pretty good test for small dev house to follow.

I have integrated in the past: admob, vungle, applovin, adbuddiz, FB and revmob. All but the first two were great disappointment. apploving and adbuddiz give ~1$ eCPM in our tests. FB get good eCPM usually but low fill rates to my surprise. 

currently admob interstitial get us ~2.5$ eCPM and vungle interstitial ~3.5$ but lower fill. 

#2

My experience by revenue (highest first):

  1. AdMob (banner and interstitial and non-rewarded video)

  2. Vungle (non-rewarded video)

  3. Chartboost. (interstitial and non-rewarded video)

iOS only.

I have little revenue from AppLovin and AdBuddiz but I send them little traffic. IIRC AdColony may not be ATS compliant so I will leave it.

I have sent higher traffic to AdBuddiz in the past but they are poor performers for me. YMMV.

Interesting thread. I do not want to spend on further plugins until the community way in with their experience - with or without actual figures.

Some more comment here:

http://iphonedevsdk.com/forum/business-legal-app-store/123392-chartboost-gives-very-low-ecpm.html

Also - I factor in ad network reliability - and I don’t mean eCPM. AdMob are here today and will be here tomorrow. Also, they pay promptly AND no daft PayPal fees etc.

I find it best to switch ad networks on and off remotely so I don’t get bogged down with a never ending cycle of PITA updates!

Be interesting to see what folks make of polls/surveys, MediaBrix and Appodeal and Unity Ads in particular…

#3

My experience. Mostly LatAm and all iOS

UnityAds - No money for the first 2K impressions. Then lower than Admob.

AdColony - The highest eCPM for me. Not ATS compliant and terrible prompts asking the player for camera and calendar access (in iOS 10)

Applovin - The highest for me after adColony. For some reason, I had to give it some time to warm up then the eCPM has increased and stayed consistent. Easy and very fast payment. 

Facebook - Either terrible fill and ok eCPM or slightly less terrible fills and lower eCPM (it depends where I put the slider).

Vungle - Lower than Applovin and AdColony. Getting info from their console is terrible. net90 pay.

Admob - Lower than AdColony and Applovin but highest fill rate. Easy and no hassle payment.

My own data. Fill rate is the second column and eCPM the last.

Data for the last 14 days.

All interstitial/video/playable games.

Max video length set to :30 seconds.

All videos skippable shortly after they start playing.

Playable games can be dismissed immediately.

Most of the traffic is LatAm and most of that is from Mexico, Colombia, and Peru.

AppLovin 80.61536223 3.47 Vungle 80.04634651 1.27 AdMob 85.6165775 1.11

Appodeal has a lot of data on what they are seeing here:

https://www.appodeal.com/reports

The response I got from AdColony when I complained about the permission requests:

Melissa White

NOV 15, 2016  |  11:53PM PST

Thank you for getting in touch with us.

 

There are no plans a this time to make any changes to the permissions needed by our network. I have however, passed on your concerns to our Product team. 

 

Here are some examples of why we require these permissions:

 

Camera and Photo Library permissions are used for campaigns that have a selfie component to their end card; camera to take the picture and Photo Library to save it. One campaign we worked with using this type of end card was the Angry Birds movie launch. This campaign had an “Angrify Yourself” selfie as part of the end card; which you can view at http://rm3.adtilt.com/2016/04/angry_birds/index.html. Note that this is a static page showing the end card layout only, as our cards rely on our SDK to provide the button functionality. But, you can see the function on your computer on the website at https://www.angrify.me/#!/. Essentially the user takes a fun selfie and the Angry Birds eye brows are added and the photo is saved; the user could take as many of these selfies as they chose and then use them off of their phone to post on social media or text to friends or whatever else they might choose.

 

The key to this functionality is that there was no requirement for the user to take a selfie. If they weren’t interested; they just closed the end card and returned to the app they were using. We will never require a user to interact with this type of function to return to their app. Additionally I want to further clarify when I mention posting the photos to social media; this is not something that our SDK or AdColony does, this is the user’s choice based on having the photo saved to their device through the Photo Library.

 

The Calendar permission allows us to create an event for the user and add it to their calendar. Again, this requires the user to opt in by clicking the button that would appear on the end card; the button will be labeled with either “Remind Me” or “Add to Calendar” for clarity to the user. When the user clicks the button, they will be taken to their calendar where they will have to approve/save the event that is created by our system. At this point they could still discard the event and return to the end card without and issue. They also have the option to not click the button at all and close the end card to return to the app that they are using. For this permission as well, we will never add to the user’s calendar without their permisison through opting in by clicking the button on the end card and approving the event. You can also view an example of an end card containing this function at http://d1e5xacotudrwg.cloudfront.net/production/builds/2016-10/22298/22b4a132-05b5-4509-abb4-140c5eb581cb/index.html. Again the “Remind Me” button will not produce a calendar event in this case as our SDK is required for this to work.

 

Please let me know if you have any further questions on this.

 

Best,

Melissa

_____________________________ 

Melissa White 

Head of Technical Support 

melissa.white@adcolony.com | http://adcolony.com/

 

 

#4

@hogletpie,

Shame about chartboost. I was about to add them to my list, but if they are spiraling down, I can save myself the trouble.

@agramonte,

Perhaps I need to give Applovin another chance for longer period. I didn’t notice that AdColony plug-in requires these permissions. I can use that in kids apps where I already ask for these due to game features, but certainly not for casual games. 

Are you using a third party tool to switch netoworks remotely ? I use my own mediation but adding a network require adding its SDK to each app so it only helps with existing networks.

#5

Just in general terms, all ad networks need to “warm up to you”.  They all have to learn about your app and it’s users trends, the number of ad impressions you’re giving over time before they start sending you better ads. 

#6

This is all from a native iOS/Swift app. I use Fyber mostly because it calculates ARPDAU by country automatically and from there it is easy for me to get LTV. Unfortunately, I don’t have a successful CoronaLabs app, but maybe 2017 will be the year I change that. I plan to use Appodeal for my corona apps. From my testing,  Appodeal has very similar functionality to Fyber (not the ARPDAU functionality, but I am being told it is coming). From the Appodeal control panel you should be able to turn on/off networks even by segments.

I have an e-mail from Fyber letting me know they take 30% unless I make some other arrangements. Those number above are after Fyber’s take.

Appodeal from all I have read and conversations with their support staff take 20%.

#7

@Rob,

From my knowledge that is mostly incorrect. Most ad networks have limited ad inventory and most end users can consume it in matter of few sessions. Out of their inventory, only a few pay nicely and this is why most players will end up seeing their ads over and over again. Only real exceptions are AdMob and FB. It’s not that networks simply throw random ads at you. They begin with highest paying ones and go down the list. Only if your audience is not responding to the ‘best sellers’, they will attempt the long tail ads, but these will not pay as handsomely, so you end up getting less than the average publisher does anyhow.

The main reason ad networks say that they need a certain amount to ‘learn’, is actually aimed of getting a larger share of the publisher pool assuming that many starting players will only test out 2-3 networks and will settle to one after few days. by then, the network will have a good chance of staying as the leading provider.

#8

@agramonte,

As a general rule, you want to keep as close to the advertiser as possible. Ad space is littered with companies often creating chains of up to 4-5 networks, each consuming a percentage of the advertiser budget. In the old web days, this fragmentation led to the introduction of ad exchanges which reduced the chain significantly, but in mobile there is still a way to go.

The main issue with aggregators is that their take is on top of the network chain (even with reduced share), so the commutative take is often higher than going direct to the networks. On the positive, you have the advantage of implementing one SDK and being able to test dozens of networks. Once you learn which networks perform better for you, its usually advisable to directly integrate with them and use your own mediation to switch priorities.

Of course, many of these networks do not share their actual take, or the quality of their counting (this alone may takes 10-15% off the actual number on either side). Result being that the best measure for us is the total amount of $ (not eCPM, not impression count etc) you get from each. Its hard to compare networks properly as in different time of the year they get different prices from advertisers and their inventory fluctuates as well.

The way I test is using google app engine, I create several versions of my back end priority file(s) and then share the app audience over several such settings during the same time period. This way you can ascertain which performs better for you. of course, this requires having significant user base or daily download count for the numbers to mean anything.

The golden goose is of course direct deals once you have enough impressions to justify the advertiser attention. I have yet to reach this level :slight_smile:

#9

My experience by revenue (highest first):

  1. AdMob (banner and interstitial and non-rewarded video)

  2. Vungle (non-rewarded video)

  3. Chartboost. (interstitial and non-rewarded video)

iOS only.

I have little revenue from AppLovin and AdBuddiz but I send them little traffic. IIRC AdColony may not be ATS compliant so I will leave it.

I have sent higher traffic to AdBuddiz in the past but they are poor performers for me. YMMV.

Interesting thread. I do not want to spend on further plugins until the community way in with their experience - with or without actual figures.

Some more comment here:

http://iphonedevsdk.com/forum/business-legal-app-store/123392-chartboost-gives-very-low-ecpm.html

Also - I factor in ad network reliability - and I don’t mean eCPM. AdMob are here today and will be here tomorrow. Also, they pay promptly AND no daft PayPal fees etc.

I find it best to switch ad networks on and off remotely so I don’t get bogged down with a never ending cycle of PITA updates!

Be interesting to see what folks make of polls/surveys, MediaBrix and Appodeal and Unity Ads in particular…

#10

My experience. Mostly LatAm and all iOS

UnityAds - No money for the first 2K impressions. Then lower than Admob.

AdColony - The highest eCPM for me. Not ATS compliant and terrible prompts asking the player for camera and calendar access (in iOS 10)

Applovin - The highest for me after adColony. For some reason, I had to give it some time to warm up then the eCPM has increased and stayed consistent. Easy and very fast payment. 

Facebook - Either terrible fill and ok eCPM or slightly less terrible fills and lower eCPM (it depends where I put the slider).

Vungle - Lower than Applovin and AdColony. Getting info from their console is terrible. net90 pay.

Admob - Lower than AdColony and Applovin but highest fill rate. Easy and no hassle payment.

My own data. Fill rate is the second column and eCPM the last.

Data for the last 14 days.

All interstitial/video/playable games.

Max video length set to :30 seconds.

All videos skippable shortly after they start playing.

Playable games can be dismissed immediately.

Most of the traffic is LatAm and most of that is from Mexico, Colombia, and Peru.

AppLovin 80.61536223 3.47 Vungle 80.04634651 1.27 AdMob 85.6165775 1.11

Appodeal has a lot of data on what they are seeing here:

https://www.appodeal.com/reports

The response I got from AdColony when I complained about the permission requests:

Melissa White

NOV 15, 2016  |  11:53PM PST

Thank you for getting in touch with us.

 

There are no plans a this time to make any changes to the permissions needed by our network. I have however, passed on your concerns to our Product team. 

 

Here are some examples of why we require these permissions:

 

Camera and Photo Library permissions are used for campaigns that have a selfie component to their end card; camera to take the picture and Photo Library to save it. One campaign we worked with using this type of end card was the Angry Birds movie launch. This campaign had an “Angrify Yourself” selfie as part of the end card; which you can view at http://rm3.adtilt.com/2016/04/angry_birds/index.html. Note that this is a static page showing the end card layout only, as our cards rely on our SDK to provide the button functionality. But, you can see the function on your computer on the website at https://www.angrify.me/#!/. Essentially the user takes a fun selfie and the Angry Birds eye brows are added and the photo is saved; the user could take as many of these selfies as they chose and then use them off of their phone to post on social media or text to friends or whatever else they might choose.

 

The key to this functionality is that there was no requirement for the user to take a selfie. If they weren’t interested; they just closed the end card and returned to the app they were using. We will never require a user to interact with this type of function to return to their app. Additionally I want to further clarify when I mention posting the photos to social media; this is not something that our SDK or AdColony does, this is the user’s choice based on having the photo saved to their device through the Photo Library.

 

The Calendar permission allows us to create an event for the user and add it to their calendar. Again, this requires the user to opt in by clicking the button that would appear on the end card; the button will be labeled with either “Remind Me” or “Add to Calendar” for clarity to the user. When the user clicks the button, they will be taken to their calendar where they will have to approve/save the event that is created by our system. At this point they could still discard the event and return to the end card without and issue. They also have the option to not click the button at all and close the end card to return to the app that they are using. For this permission as well, we will never add to the user’s calendar without their permisison through opting in by clicking the button on the end card and approving the event. You can also view an example of an end card containing this function at http://d1e5xacotudrwg.cloudfront.net/production/builds/2016-10/22298/22b4a132-05b5-4509-abb4-140c5eb581cb/index.html. Again the “Remind Me” button will not produce a calendar event in this case as our SDK is required for this to work.

 

Please let me know if you have any further questions on this.

 

Best,

Melissa

_____________________________ 

Melissa White 

Head of Technical Support 

melissa.white@adcolony.com | http://adcolony.com/

 

 

#11

@hogletpie,

Shame about chartboost. I was about to add them to my list, but if they are spiraling down, I can save myself the trouble.

@agramonte,

Perhaps I need to give Applovin another chance for longer period. I didn’t notice that AdColony plug-in requires these permissions. I can use that in kids apps where I already ask for these due to game features, but certainly not for casual games. 

Are you using a third party tool to switch netoworks remotely ? I use my own mediation but adding a network require adding its SDK to each app so it only helps with existing networks.

#12

Just in general terms, all ad networks need to “warm up to you”.  They all have to learn about your app and it’s users trends, the number of ad impressions you’re giving over time before they start sending you better ads. 

#13

This is all from a native iOS/Swift app. I use Fyber mostly because it calculates ARPDAU by country automatically and from there it is easy for me to get LTV. Unfortunately, I don’t have a successful CoronaLabs app, but maybe 2017 will be the year I change that. I plan to use Appodeal for my corona apps. From my testing,  Appodeal has very similar functionality to Fyber (not the ARPDAU functionality, but I am being told it is coming). From the Appodeal control panel you should be able to turn on/off networks even by segments.

I have an e-mail from Fyber letting me know they take 30% unless I make some other arrangements. Those number above are after Fyber’s take.

Appodeal from all I have read and conversations with their support staff take 20%.

#14

@Rob,

From my knowledge that is mostly incorrect. Most ad networks have limited ad inventory and most end users can consume it in matter of few sessions. Out of their inventory, only a few pay nicely and this is why most players will end up seeing their ads over and over again. Only real exceptions are AdMob and FB. It’s not that networks simply throw random ads at you. They begin with highest paying ones and go down the list. Only if your audience is not responding to the ‘best sellers’, they will attempt the long tail ads, but these will not pay as handsomely, so you end up getting less than the average publisher does anyhow.

The main reason ad networks say that they need a certain amount to ‘learn’, is actually aimed of getting a larger share of the publisher pool assuming that many starting players will only test out 2-3 networks and will settle to one after few days. by then, the network will have a good chance of staying as the leading provider.

#15

@agramonte,

As a general rule, you want to keep as close to the advertiser as possible. Ad space is littered with companies often creating chains of up to 4-5 networks, each consuming a percentage of the advertiser budget. In the old web days, this fragmentation led to the introduction of ad exchanges which reduced the chain significantly, but in mobile there is still a way to go.

The main issue with aggregators is that their take is on top of the network chain (even with reduced share), so the commutative take is often higher than going direct to the networks. On the positive, you have the advantage of implementing one SDK and being able to test dozens of networks. Once you learn which networks perform better for you, its usually advisable to directly integrate with them and use your own mediation to switch priorities.

Of course, many of these networks do not share their actual take, or the quality of their counting (this alone may takes 10-15% off the actual number on either side). Result being that the best measure for us is the total amount of $ (not eCPM, not impression count etc) you get from each. Its hard to compare networks properly as in different time of the year they get different prices from advertisers and their inventory fluctuates as well.

The way I test is using google app engine, I create several versions of my back end priority file(s) and then share the app audience over several such settings during the same time period. This way you can ascertain which performs better for you. of course, this requires having significant user base or daily download count for the numbers to mean anything.

The golden goose is of course direct deals once you have enough impressions to justify the advertiser attention. I have yet to reach this level :slight_smile: