Josh Chandley (Chief Operating Officer at WildCard) shares the studio's current mobile marketing strategy and discusses with Nebo Radovic (UA Platform at N3TWORK) and Jon Lau (Director of Growth at Weee!) on how to take it to the next level.
A way to test creatives at a lower cost is to start in India with a small budget to filter them out first, then move to Canada and deploying the best-performing ones in the US.
For games that have a strong organic demand, it makes sense to pay attention to ASO and also to potentially try preloads on devices as well as try low cost/quality channels. Examples: Gin, Rummy, Solitaire, Tetris.
For games like card games that have organic interest it make sense to work with interests and keyword targeting because there is an actual "card games interest" that you can target.
When you're an ad-monetized game the LTV curve of AEO/VO is really different from the LTV curve of MAI/rewarded video players. So you need to spend enough money to understand how that LTV curve looks like for each then if it hits your DX payback target you invest as much money as you can.
Showing more interstitials to users might not result in a higher net LTV long term but if it changes the shape of your ARPU curve and reduces your payback window it allows you to spend more aggressively.
You need to have user-level ad revenue to understand what the value of each user is and the overall impact on returns (not just campaigns) because you are able to monetize better the users you get through AEO/VO via higher ad CPMs. You have to know to which extent you can monetize them better and if it offsets the higher acquisition costs so you can decide where to invest (between AEO, VO, MAI, etc.). However it's also risky to bet on acquiring these "whales" you get through AEO/VO because you might also all of a sudden not be able to retain these high-value users which can make your CPM drop.
Diversify if you're scaling up an ad-driven game because there is a hidden risk in relying on AEO/VO campaigns, especially if it's optimized for purchase (however it could work well for ad views or engagement). You have ways to get the same user through other mediums.
It's important to have ad LTV curve and IAP LTV curve separate to understand the impact of each on payback. Then you want granularity by campaign, optimization model (AEO/VO, AEO for ad views, rewarded videos, etc.)
If your LTV curves are jumping around, one thing you might want to consider is testing your scale of spend against your ROAS target. Example: looking at D7 ROAS against daily spend to see if you have a knee bend. If you're unsure about spend you can "sit below the bend".
Although it can depend on the game, one creative format (static, video, etc.) does not necessarily drive higher ROAS or bring you a better user than another. But the lifetime of a creative format can change, with video performance lasting longer for example.
Focus on diversifying your creative mix so you can elongate the overall lifetime of your creatives. Try to maintain a higher average IPM of your overall creative mix rather than chasing finding one or two creatives that are going to be hits.
Focus more on creating "micro-videos". Shoot and create multiple 3-5 seconds videos (product features, why it's fun, "group play", etc.) and create 15 different combinations from these videos. It allows you to do very little video production and have a bigger outcome. Example (e-commerce grocery business): instead of doing a video focused on 1 product for 15 seconds, use several shots from different products to mix and match and create a lot of different videos like "just snacks", "snacks and desserts", etc.
If a game like Cribbage is relevant to specific geos it might be interesting to group campaigns together to maintain the cost and make it easier for Facebook to learn. Sometimes when companies scale they create a lot of campaigns that end up stuck in the learning phase.
Because AEO and VO are optimizing for different purchasing behaviors, you might be able to target ("naturally" since you can't control that) different audiences for each type of campaign and leverage this to either go hard on pushing IAP/pricing vs. monetizing.
Creatives is the biggest lever. Never fall in love with your idea or your hypothesis, let the data speak. Trust the process, not a specific ad tech.
A way to test creatives at a lower cost is to start in India with a small budget to filter them out first, then move to Canada and deploying the best-performing ones in the US.
For games that have a strong organic demand, it makes sense to pay attention to ASO and also to potentially try preloads on devices as well as try low cost/quality channels. Examples: Gin, Rummy, Solitaire, Tetris.
For games like card games that have organic interest it make sense to work with interests and keyword targeting because there is an actual "card games interest" that you can target.
When you're an ad-monetized game the LTV curve of AEO/VO is really different from the LTV curve of MAI/rewarded video players. So you need to spend enough money to understand how that LTV curve looks like for each then if it hits your DX payback target you invest as much money as you can.
Showing more interstitials to users might not result in a higher net LTV long term but if it changes the shape of your ARPU curve and reduces your payback window it allows you to spend more aggressively.
You need to have user-level ad revenue to understand what the value of each user is and the overall impact on returns (not just campaigns) because you are able to monetize better the users you get through AEO/VO via higher ad CPMs. You have to know to which extent you can monetize them better and if it offsets the higher acquisition costs so you can decide where to invest (between AEO, VO, MAI, etc.). However it's also risky to bet on acquiring these "whales" you get through AEO/VO because you might also all of a sudden not be able to retain these high-value users which can make your CPM drop.
Diversify if you're scaling up an ad-driven game because there is a hidden risk in relying on AEO/VO campaigns, especially if it's optimized for purchase (however it could work well for ad views or engagement). You have ways to get the same user through other mediums.
It's important to have ad LTV curve and IAP LTV curve separate to understand the impact of each on payback. Then you want granularity by campaign, optimization model (AEO/VO, AEO for ad views, rewarded videos, etc.)
If your LTV curves are jumping around, one thing you might want to consider is testing your scale of spend against your ROAS target. Example: looking at D7 ROAS against daily spend to see if you have a knee bend. If you're unsure about spend you can "sit below the bend".
Although it can depend on the game, one creative format (static, video, etc.) does not necessarily drive higher ROAS or bring you a better user than another. But the lifetime of a creative format can change, with video performance lasting longer for example.
Focus on diversifying your creative mix so you can elongate the overall lifetime of your creatives. Try to maintain a higher average IPM of your overall creative mix rather than chasing finding one or two creatives that are going to be hits.
Focus more on creating "micro-videos". Shoot and create multiple 3-5 seconds videos (product features, why it's fun, "group play", etc.) and create 15 different combinations from these videos. It allows you to do very little video production and have a bigger outcome. Example (e-commerce grocery business): instead of doing a video focused on 1 product for 15 seconds, use several shots from different products to mix and match and create a lot of different videos like "just snacks", "snacks and desserts", etc.
If a game like Cribbage is relevant to specific geos it might be interesting to group campaigns together to maintain the cost and make it easier for Facebook to learn. Sometimes when companies scale they create a lot of campaigns that end up stuck in the learning phase.
Because AEO and VO are optimizing for different purchasing behaviors, you might be able to target ("naturally" since you can't control that) different audiences for each type of campaign and leverage this to either go hard on pushing IAP/pricing vs. monetizing.
Creatives is the biggest lever. Never fall in love with your idea or your hypothesis, let the data speak. Trust the process, not a specific ad tech.
A way to test creatives at a lower cost is to start in India with a small budget to filter them out first, then move to Canada and deploying the best-performing ones in the US.
For games that have a strong organic demand, it makes sense to pay attention to ASO and also to potentially try preloads on devices as well as try low cost/quality channels. Examples: Gin, Rummy, Solitaire, Tetris.
For games like card games that have organic interest it make sense to work with interests and keyword targeting because there is an actual "card games interest" that you can target.
When you're an ad-monetized game the LTV curve of AEO/VO is really different from the LTV curve of MAI/rewarded video players. So you need to spend enough money to understand how that LTV curve looks like for each then if it hits your DX payback target you invest as much money as you can.
Showing more interstitials to users might not result in a higher net LTV long term but if it changes the shape of your ARPU curve and reduces your payback window it allows you to spend more aggressively.
You need to have user-level ad revenue to understand what the value of each user is and the overall impact on returns (not just campaigns) because you are able to monetize better the users you get through AEO/VO via higher ad CPMs. You have to know to which extent you can monetize them better and if it offsets the higher acquisition costs so you can decide where to invest (between AEO, VO, MAI, etc.). However it's also risky to bet on acquiring these "whales" you get through AEO/VO because you might also all of a sudden not be able to retain these high-value users which can make your CPM drop.
Diversify if you're scaling up an ad-driven game because there is a hidden risk in relying on AEO/VO campaigns, especially if it's optimized for purchase (however it could work well for ad views or engagement). You have ways to get the same user through other mediums.
It's important to have ad LTV curve and IAP LTV curve separate to understand the impact of each on payback. Then you want granularity by campaign, optimization model (AEO/VO, AEO for ad views, rewarded videos, etc.)
If your LTV curves are jumping around, one thing you might want to consider is testing your scale of spend against your ROAS target. Example: looking at D7 ROAS against daily spend to see if you have a knee bend. If you're unsure about spend you can "sit below the bend".
Although it can depend on the game, one creative format (static, video, etc.) does not necessarily drive higher ROAS or bring you a better user than another. But the lifetime of a creative format can change, with video performance lasting longer for example.
Focus on diversifying your creative mix so you can elongate the overall lifetime of your creatives. Try to maintain a higher average IPM of your overall creative mix rather than chasing finding one or two creatives that are going to be hits.
Focus more on creating "micro-videos". Shoot and create multiple 3-5 seconds videos (product features, why it's fun, "group play", etc.) and create 15 different combinations from these videos. It allows you to do very little video production and have a bigger outcome. Example (e-commerce grocery business): instead of doing a video focused on 1 product for 15 seconds, use several shots from different products to mix and match and create a lot of different videos like "just snacks", "snacks and desserts", etc.
If a game like Cribbage is relevant to specific geos it might be interesting to group campaigns together to maintain the cost and make it easier for Facebook to learn. Sometimes when companies scale they create a lot of campaigns that end up stuck in the learning phase.
Because AEO and VO are optimizing for different purchasing behaviors, you might be able to target ("naturally" since you can't control that) different audiences for each type of campaign and leverage this to either go hard on pushing IAP/pricing vs. monetizing.
Creatives is the biggest lever. Never fall in love with your idea or your hypothesis, let the data speak. Trust the process, not a specific ad tech.
Notes for this resource are currently being transferred and will be available soon.
Wildcards is re-imagining card games for mobile.
Started with running just a couple hundreds of dollars on Facebook. Then scaled gradually.
Primarily on FB (main focus), growing fast on Google and some networks (Vungle, Unity Ads, AppLovin).
On Facebook:
They create a new audience list every month to try and find something that works.
Creative side: working with Consumer Acquisition who produces about 10 videos per month and runs their UA.
[💎 @04:44] Consumer Acquisition has a testing process for creatives where they start in India with a small budget to filter first, then move to Canada and if it works they are used in the US.
Here is how they are determining budgets: they want to breakeven between D90-120 on Ad Spend and they set a D7 target for that (differently for Mobile App Installs and VO). Then Consumer Acquisition uses D7 to determine their D1/D3 targets.
Card games are specific because people can actually be searching for them, which is not necessarily the case for other titles.
[💎 @06:45] For games that have a strong organic demand, it makes sense to pay attention to ASO and also to potentially try preloads on devices as well as try low cost/quality channels. Examples: Gin, Rummy, Solitaire, Tetris.
These games tend to be low CAC.
[💎 @07:47] For games like card games that do have organic interest it does make sense to work with interests and keyword targeting because there is an actual "card games interest" that you can target.
How to think about budget for other networks?
[💎 @08:57] When you're an ad-monetized game the LTV curve of AEO/VO is really different from the LTV curve of MAI/rewarded video players. So you need to spend enough money to understand how that LTV curve looks like for each then if it hits your DX payback target you invest as much money as you can.
D7 ROAS significantly higher in AEO/VO then after D30 MAI is taking over (better than AEO/VO) because of the low costs (the ad monetization "takes over") but then it grows steadily but slowly. So the acquisition cost for an install makes a huge difference.
Game first launched in 2016. Only been scaling in the last 6-12 months after putting in place the back-end tech allowing to run A/B tests non-stop and measuring ARPU curve for these tests at impression level granularity.
Put a lot of effort into measuring ARPU curve by source and by country for D180, which allowed to buy much more aggressively. Then they started growing by running the A/B tests and 1 (more interstitial) boosted D1 ARPU by 100% and D30 by 30%. That would time they were buying on historical data even though each test is boosting ARPU.
So now they try to not only look at historical data but also project where the ARPU curve is going to be.
[💎 @14:36] Showing more interstitials to users might not result in a higher net LTV long term but if it changes the shape of your ARPU curve and reduces your payback window it allows you to spend more aggressively.
Audience matches with slots and match games.
When running AEO/VO the upfront costs are very expensive because there are whales. With ad-driven games and lower ARPU you're always worried about cost/CPI and basing performance on paid only is going to make it difficult to achieve payback with AEO/VO campaigns.
If you're ok with blended outlook (with organics) it's also dangerous because you need to make sure you have enough organics.
Look closer at how you can utilize interest targeting and ASO.
What's worked great for Wildcard in terms of scalability is low-cost Google search and Apple Search Ads. Other things didn't have that kind of scale.
Use Mopub for mediation and they have an impression-level revenue tool.
You need different LTV curves AEO/VO curves and MAI/rewarded.
[💎 @20:54] You need to have user-level ad revenue to understand what's the value of each user and the overall impact on returns (not just campaigns) because you are able to monetize better the users you get through AEO/VO via higher ad CPMs. You have to know to which extent you can monetize them better and if it offsets the higher acquisition costs so you can decide where to invest (between AEO, VO, MAI, etc.). However it's also risky to bet on acquiring these "whales" you get through AEO/VO because you might also all of a sudden not be able to retain these high-value users which can make your CPM drop.
LTV of a card game happens over a longer time than a midcore game and during that time there can be a lot of things can happen: competition, seasonality, etc.
Using VO/AEO (left: first days look good then LTV curve flattens and you don't reach payback)) vs. using lower CPI strategy for ad-driven businesses (right: lower cost and you might get payback faster)
[💎 @29:20] It's important to have ad LTV curve and IAP LTV curve separate to understand the impact of each on payback. Then you want granularity by campaign, optimization model (AEO/VO, AEO for ad views, rewarded videos, etc.).
[💎 @29:48] If your LTV curves are jumping around, one thing you might want to consider is testing your scale of spend against your ROAS target. Example: looking at D7 ROAS against daily spend to see if you have a knee bend. If you're unsure about spend you can "sit below the bend".
More info by Joseph on this after reaching out:
So basically, my point was that for many games there is a somewhat consistent scale to spend before it "knee bends". In other words, if you were to measure your daily spend against your Dx target. Let's say it's a D3 or D7 target. then basically just create a table of daily spend and then what the D7 ROAS is against that spend. What you will often see is a degradation in ROAS that quickly goes down after a certain point... For the last game I worked on it was something like $20K -$25K per day on FB for example.
So I was just saying, if Josh was making a bunch of changes that was screwing up the data, to try and sit under that daily spend.
Current approach:
What they've seen is that videos stale slower but no evidence that videos will outperform significantly.
[💎 @32:39] Although it can depend on the game, one creative format (static, video, etc.) does not necessarily drive higher ROAS or bring you a better user. But the lifetime can change with video performance lasting longer for example.
It's easier to create statics so instead of trying to find a breakout video it might be better to diversify the creative mix.
[💎 @33:10] Focus on diversifying your creative mix so you can elongate the overall lifetime of your creatives. Try to maintain a higher average IPM of your overall creative mix rather than one or two creatives that are going to be hits.
Article from Nebo on mainting healthy IPM
It's tough to know what's going to be a hit and what's not. Consumer Acquisition once produced a one-year hit for Playstudios even though everybody thought it would not work.
Instead of focusing on cranking a lot of videos and hoping one of them is a winner, focus on making sure you're maintaining the IPM rate you have across your portfolio mix.
Once you really scale video has to be part of your creative mix: different dimensions, rewarded video specific sizes, etc.
The audience playing Cribbage is older and more responsive to ads than younger audiences. That's probably one reason statics are working.
[💎 @37:00] Focus more on creating "micro-videos". Shoot and create multiple 3-5 seconds videos (product features, why it's fun, "group play", etc.) and create 15 different combinations from these videos. It allows you to do very little video production and have a bigger outcome.
Example (e-commerce grocery business): instead of doing a video focused on 1 product for 15 seconds, use several shots from different products to mix and match and create a lot of different videos like "just snacks", "snacks and desserts", etc.
It puts less pressure on the creative team to produce new creatives often and allows to refresh campaigns more often.
After 100-250k monthly spend range, what stops working?
You want to be conscious of your costs. But when you scale you have more information and data that help you make better decisions.
[💎 @41:27] If a game like Cribbage is relevant to specific geos it might be interesting to group campaigns together to maintain the cost and make it easier for Facebook to learn. Sometimes when companies scale they create a lot of campaigns that end up stuck in the learning phase.
When you scale you also need to make sure creative fatigue is not a problem.
When you're scaling, stand firm on your payback.
How will that impact programmatic campaigns? Nobody has a good idea of what's going to happen but we know that:
If you make changes on the product side, does it put you back in learning phase? Not really, this happens when you change something significantly on the campaign side. However if you front load most of the monetization it can help algos optimize better.
Possible difference in the way Facebook targets audiences for AEO and VO campaigns. Example: e-commerce using VO to look for buyers that buy in the first few days and pay a lot, similar to gaming looking for whales. But if you're doing AEO it might target different people that make purchases towards the end of the 7 day lookback window.
[💎 @48:07] Because AEO and VO are optimizing for different purchasing behaviors, you might be able to target ("naturally" since you can't control that) different audiences for each type of campaign and leverage this to either go hard on pushing IAP/pricing vs. monetizing.
Josh mentioned GameTune, a tool using machine learning to adjust your game in real-time and deliver the optimal experience to your players.
[💎 @50:36] Creatives is the biggest lever. Never fall in love with your idea or your hypothesis, let the data speak. Trust the process, not a specific ad tech.
Wildcards is re-imagining card games for mobile.
Started with running just a couple hundreds of dollars on Facebook. Then scaled gradually.
Primarily on FB (main focus), growing fast on Google and some networks (Vungle, Unity Ads, AppLovin).
On Facebook:
They create a new audience list every month to try and find something that works.
Creative side: working with Consumer Acquisition who produces about 10 videos per month and runs their UA.
[💎 @04:44] Consumer Acquisition has a testing process for creatives where they start in India with a small budget to filter first, then move to Canada and if it works they are used in the US.
Here is how they are determining budgets: they want to breakeven between D90-120 on Ad Spend and they set a D7 target for that (differently for Mobile App Installs and VO). Then Consumer Acquisition uses D7 to determine their D1/D3 targets.
Card games are specific because people can actually be searching for them, which is not necessarily the case for other titles.
[💎 @06:45] For games that have a strong organic demand, it makes sense to pay attention to ASO and also to potentially try preloads on devices as well as try low cost/quality channels. Examples: Gin, Rummy, Solitaire, Tetris.
These games tend to be low CAC.
[💎 @07:47] For games like card games that do have organic interest it does make sense to work with interests and keyword targeting because there is an actual "card games interest" that you can target.
How to think about budget for other networks?
[💎 @08:57] When you're an ad-monetized game the LTV curve of AEO/VO is really different from the LTV curve of MAI/rewarded video players. So you need to spend enough money to understand how that LTV curve looks like for each then if it hits your DX payback target you invest as much money as you can.
D7 ROAS significantly higher in AEO/VO then after D30 MAI is taking over (better than AEO/VO) because of the low costs (the ad monetization "takes over") but then it grows steadily but slowly. So the acquisition cost for an install makes a huge difference.
Game first launched in 2016. Only been scaling in the last 6-12 months after putting in place the back-end tech allowing to run A/B tests non-stop and measuring ARPU curve for these tests at impression level granularity.
Put a lot of effort into measuring ARPU curve by source and by country for D180, which allowed to buy much more aggressively. Then they started growing by running the A/B tests and 1 (more interstitial) boosted D1 ARPU by 100% and D30 by 30%. That would time they were buying on historical data even though each test is boosting ARPU.
So now they try to not only look at historical data but also project where the ARPU curve is going to be.
[💎 @14:36] Showing more interstitials to users might not result in a higher net LTV long term but if it changes the shape of your ARPU curve and reduces your payback window it allows you to spend more aggressively.
Audience matches with slots and match games.
When running AEO/VO the upfront costs are very expensive because there are whales. With ad-driven games and lower ARPU you're always worried about cost/CPI and basing performance on paid only is going to make it difficult to achieve payback with AEO/VO campaigns.
If you're ok with blended outlook (with organics) it's also dangerous because you need to make sure you have enough organics.
Look closer at how you can utilize interest targeting and ASO.
What's worked great for Wildcard in terms of scalability is low-cost Google search and Apple Search Ads. Other things didn't have that kind of scale.
Use Mopub for mediation and they have an impression-level revenue tool.
You need different LTV curves AEO/VO curves and MAI/rewarded.
[💎 @20:54] You need to have user-level ad revenue to understand what's the value of each user and the overall impact on returns (not just campaigns) because you are able to monetize better the users you get through AEO/VO via higher ad CPMs. You have to know to which extent you can monetize them better and if it offsets the higher acquisition costs so you can decide where to invest (between AEO, VO, MAI, etc.). However it's also risky to bet on acquiring these "whales" you get through AEO/VO because you might also all of a sudden not be able to retain these high-value users which can make your CPM drop.
LTV of a card game happens over a longer time than a midcore game and during that time there can be a lot of things can happen: competition, seasonality, etc.
Using VO/AEO (left: first days look good then LTV curve flattens and you don't reach payback)) vs. using lower CPI strategy for ad-driven businesses (right: lower cost and you might get payback faster)
[💎 @29:20] It's important to have ad LTV curve and IAP LTV curve separate to understand the impact of each on payback. Then you want granularity by campaign, optimization model (AEO/VO, AEO for ad views, rewarded videos, etc.).
[💎 @29:48] If your LTV curves are jumping around, one thing you might want to consider is testing your scale of spend against your ROAS target. Example: looking at D7 ROAS against daily spend to see if you have a knee bend. If you're unsure about spend you can "sit below the bend".
More info by Joseph on this after reaching out:
So basically, my point was that for many games there is a somewhat consistent scale to spend before it "knee bends". In other words, if you were to measure your daily spend against your Dx target. Let's say it's a D3 or D7 target. then basically just create a table of daily spend and then what the D7 ROAS is against that spend. What you will often see is a degradation in ROAS that quickly goes down after a certain point... For the last game I worked on it was something like $20K -$25K per day on FB for example.
So I was just saying, if Josh was making a bunch of changes that was screwing up the data, to try and sit under that daily spend.
Current approach:
What they've seen is that videos stale slower but no evidence that videos will outperform significantly.
[💎 @32:39] Although it can depend on the game, one creative format (static, video, etc.) does not necessarily drive higher ROAS or bring you a better user. But the lifetime can change with video performance lasting longer for example.
It's easier to create statics so instead of trying to find a breakout video it might be better to diversify the creative mix.
[💎 @33:10] Focus on diversifying your creative mix so you can elongate the overall lifetime of your creatives. Try to maintain a higher average IPM of your overall creative mix rather than one or two creatives that are going to be hits.
Article from Nebo on mainting healthy IPM
It's tough to know what's going to be a hit and what's not. Consumer Acquisition once produced a one-year hit for Playstudios even though everybody thought it would not work.
Instead of focusing on cranking a lot of videos and hoping one of them is a winner, focus on making sure you're maintaining the IPM rate you have across your portfolio mix.
Once you really scale video has to be part of your creative mix: different dimensions, rewarded video specific sizes, etc.
The audience playing Cribbage is older and more responsive to ads than younger audiences. That's probably one reason statics are working.
[💎 @37:00] Focus more on creating "micro-videos". Shoot and create multiple 3-5 seconds videos (product features, why it's fun, "group play", etc.) and create 15 different combinations from these videos. It allows you to do very little video production and have a bigger outcome.
Example (e-commerce grocery business): instead of doing a video focused on 1 product for 15 seconds, use several shots from different products to mix and match and create a lot of different videos like "just snacks", "snacks and desserts", etc.
It puts less pressure on the creative team to produce new creatives often and allows to refresh campaigns more often.
After 100-250k monthly spend range, what stops working?
You want to be conscious of your costs. But when you scale you have more information and data that help you make better decisions.
[💎 @41:27] If a game like Cribbage is relevant to specific geos it might be interesting to group campaigns together to maintain the cost and make it easier for Facebook to learn. Sometimes when companies scale they create a lot of campaigns that end up stuck in the learning phase.
When you scale you also need to make sure creative fatigue is not a problem.
When you're scaling, stand firm on your payback.
How will that impact programmatic campaigns? Nobody has a good idea of what's going to happen but we know that:
If you make changes on the product side, does it put you back in learning phase? Not really, this happens when you change something significantly on the campaign side. However if you front load most of the monetization it can help algos optimize better.
Possible difference in the way Facebook targets audiences for AEO and VO campaigns. Example: e-commerce using VO to look for buyers that buy in the first few days and pay a lot, similar to gaming looking for whales. But if you're doing AEO it might target different people that make purchases towards the end of the 7 day lookback window.
[💎 @48:07] Because AEO and VO are optimizing for different purchasing behaviors, you might be able to target ("naturally" since you can't control that) different audiences for each type of campaign and leverage this to either go hard on pushing IAP/pricing vs. monetizing.
Josh mentioned GameTune, a tool using machine learning to adjust your game in real-time and deliver the optimal experience to your players.
[💎 @50:36] Creatives is the biggest lever. Never fall in love with your idea or your hypothesis, let the data speak. Trust the process, not a specific ad tech.
Wildcards is re-imagining card games for mobile.
Started with running just a couple hundreds of dollars on Facebook. Then scaled gradually.
Primarily on FB (main focus), growing fast on Google and some networks (Vungle, Unity Ads, AppLovin).
On Facebook:
They create a new audience list every month to try and find something that works.
Creative side: working with Consumer Acquisition who produces about 10 videos per month and runs their UA.
[💎 @04:44] Consumer Acquisition has a testing process for creatives where they start in India with a small budget to filter first, then move to Canada and if it works they are used in the US.
Here is how they are determining budgets: they want to breakeven between D90-120 on Ad Spend and they set a D7 target for that (differently for Mobile App Installs and VO). Then Consumer Acquisition uses D7 to determine their D1/D3 targets.
Card games are specific because people can actually be searching for them, which is not necessarily the case for other titles.
[💎 @06:45] For games that have a strong organic demand, it makes sense to pay attention to ASO and also to potentially try preloads on devices as well as try low cost/quality channels. Examples: Gin, Rummy, Solitaire, Tetris.
These games tend to be low CAC.
[💎 @07:47] For games like card games that do have organic interest it does make sense to work with interests and keyword targeting because there is an actual "card games interest" that you can target.
How to think about budget for other networks?
[💎 @08:57] When you're an ad-monetized game the LTV curve of AEO/VO is really different from the LTV curve of MAI/rewarded video players. So you need to spend enough money to understand how that LTV curve looks like for each then if it hits your DX payback target you invest as much money as you can.
D7 ROAS significantly higher in AEO/VO then after D30 MAI is taking over (better than AEO/VO) because of the low costs (the ad monetization "takes over") but then it grows steadily but slowly. So the acquisition cost for an install makes a huge difference.
Game first launched in 2016. Only been scaling in the last 6-12 months after putting in place the back-end tech allowing to run A/B tests non-stop and measuring ARPU curve for these tests at impression level granularity.
Put a lot of effort into measuring ARPU curve by source and by country for D180, which allowed to buy much more aggressively. Then they started growing by running the A/B tests and 1 (more interstitial) boosted D1 ARPU by 100% and D30 by 30%. That would time they were buying on historical data even though each test is boosting ARPU.
So now they try to not only look at historical data but also project where the ARPU curve is going to be.
[💎 @14:36] Showing more interstitials to users might not result in a higher net LTV long term but if it changes the shape of your ARPU curve and reduces your payback window it allows you to spend more aggressively.
Audience matches with slots and match games.
When running AEO/VO the upfront costs are very expensive because there are whales. With ad-driven games and lower ARPU you're always worried about cost/CPI and basing performance on paid only is going to make it difficult to achieve payback with AEO/VO campaigns.
If you're ok with blended outlook (with organics) it's also dangerous because you need to make sure you have enough organics.
Look closer at how you can utilize interest targeting and ASO.
What's worked great for Wildcard in terms of scalability is low-cost Google search and Apple Search Ads. Other things didn't have that kind of scale.
Use Mopub for mediation and they have an impression-level revenue tool.
You need different LTV curves AEO/VO curves and MAI/rewarded.
[💎 @20:54] You need to have user-level ad revenue to understand what's the value of each user and the overall impact on returns (not just campaigns) because you are able to monetize better the users you get through AEO/VO via higher ad CPMs. You have to know to which extent you can monetize them better and if it offsets the higher acquisition costs so you can decide where to invest (between AEO, VO, MAI, etc.). However it's also risky to bet on acquiring these "whales" you get through AEO/VO because you might also all of a sudden not be able to retain these high-value users which can make your CPM drop.
LTV of a card game happens over a longer time than a midcore game and during that time there can be a lot of things can happen: competition, seasonality, etc.
Using VO/AEO (left: first days look good then LTV curve flattens and you don't reach payback)) vs. using lower CPI strategy for ad-driven businesses (right: lower cost and you might get payback faster)
[💎 @29:20] It's important to have ad LTV curve and IAP LTV curve separate to understand the impact of each on payback. Then you want granularity by campaign, optimization model (AEO/VO, AEO for ad views, rewarded videos, etc.).
[💎 @29:48] If your LTV curves are jumping around, one thing you might want to consider is testing your scale of spend against your ROAS target. Example: looking at D7 ROAS against daily spend to see if you have a knee bend. If you're unsure about spend you can "sit below the bend".
More info by Joseph on this after reaching out:
So basically, my point was that for many games there is a somewhat consistent scale to spend before it "knee bends". In other words, if you were to measure your daily spend against your Dx target. Let's say it's a D3 or D7 target. then basically just create a table of daily spend and then what the D7 ROAS is against that spend. What you will often see is a degradation in ROAS that quickly goes down after a certain point... For the last game I worked on it was something like $20K -$25K per day on FB for example.
So I was just saying, if Josh was making a bunch of changes that was screwing up the data, to try and sit under that daily spend.
Current approach:
What they've seen is that videos stale slower but no evidence that videos will outperform significantly.
[💎 @32:39] Although it can depend on the game, one creative format (static, video, etc.) does not necessarily drive higher ROAS or bring you a better user. But the lifetime can change with video performance lasting longer for example.
It's easier to create statics so instead of trying to find a breakout video it might be better to diversify the creative mix.
[💎 @33:10] Focus on diversifying your creative mix so you can elongate the overall lifetime of your creatives. Try to maintain a higher average IPM of your overall creative mix rather than one or two creatives that are going to be hits.
Article from Nebo on mainting healthy IPM
It's tough to know what's going to be a hit and what's not. Consumer Acquisition once produced a one-year hit for Playstudios even though everybody thought it would not work.
Instead of focusing on cranking a lot of videos and hoping one of them is a winner, focus on making sure you're maintaining the IPM rate you have across your portfolio mix.
Once you really scale video has to be part of your creative mix: different dimensions, rewarded video specific sizes, etc.
The audience playing Cribbage is older and more responsive to ads than younger audiences. That's probably one reason statics are working.
[💎 @37:00] Focus more on creating "micro-videos". Shoot and create multiple 3-5 seconds videos (product features, why it's fun, "group play", etc.) and create 15 different combinations from these videos. It allows you to do very little video production and have a bigger outcome.
Example (e-commerce grocery business): instead of doing a video focused on 1 product for 15 seconds, use several shots from different products to mix and match and create a lot of different videos like "just snacks", "snacks and desserts", etc.
It puts less pressure on the creative team to produce new creatives often and allows to refresh campaigns more often.
After 100-250k monthly spend range, what stops working?
You want to be conscious of your costs. But when you scale you have more information and data that help you make better decisions.
[💎 @41:27] If a game like Cribbage is relevant to specific geos it might be interesting to group campaigns together to maintain the cost and make it easier for Facebook to learn. Sometimes when companies scale they create a lot of campaigns that end up stuck in the learning phase.
When you scale you also need to make sure creative fatigue is not a problem.
When you're scaling, stand firm on your payback.
How will that impact programmatic campaigns? Nobody has a good idea of what's going to happen but we know that:
If you make changes on the product side, does it put you back in learning phase? Not really, this happens when you change something significantly on the campaign side. However if you front load most of the monetization it can help algos optimize better.
Possible difference in the way Facebook targets audiences for AEO and VO campaigns. Example: e-commerce using VO to look for buyers that buy in the first few days and pay a lot, similar to gaming looking for whales. But if you're doing AEO it might target different people that make purchases towards the end of the 7 day lookback window.
[💎 @48:07] Because AEO and VO are optimizing for different purchasing behaviors, you might be able to target ("naturally" since you can't control that) different audiences for each type of campaign and leverage this to either go hard on pushing IAP/pricing vs. monetizing.
Josh mentioned GameTune, a tool using machine learning to adjust your game in real-time and deliver the optimal experience to your players.
[💎 @50:36] Creatives is the biggest lever. Never fall in love with your idea or your hypothesis, let the data speak. Trust the process, not a specific ad tech.
