Advanced ad strategies that drive revenue and boost engagement

[Music] I'm John Walsh and I think there's still people here oh yeah it's a couple of good-looking guys right there tell you a little bit about myself it's the wrong font of course but not a big deal found it he's powered in 2009 fortunate in the early days of mobile games to have some number-one game titles for you to play game titles and built at a publishing platform around that one key component of the publishing platform we built was a mediation platform and that's what I'm going to talk to you guys a little bit about today and what we've learned and about a year ago this time officially joined up with up site brought my company fees powered combined with upside to create one unified super powered SDK but I'm not going to talk about that today I'm going to talk about ad monetization and three easy steps to literally making four times more ad

revenue first we'll look back at the evolution of mobile ads you may remember if you're as old as me back in 2009-2010 banner ads for the big thing we evolved a static interstitial our good pal to chartboost really innovated that nice job Nate and then we evolved to video interstitial and ultimately rewarded video I'm here to talk about the next step advanced rewarded video and talk to you about how you can make a lot more money by doing some basic but pretty interesting things so we're going to use some stats from an actual app but kind of rounded the numbers to make the math a little easier to digest but this is a base case app that had 200 thousand roughly daily active users delivering 200 thousand impressions ad impressions rewarded video ad impressions specifically per day generating about a ten dollar global eCPM and one of the rules that this publisher had in place was that they were capping the number of rewarded videos a user could see every day at 30 seems like a lot but sort of

an arbitrary cap that they put in place first assumption is 200 thousand daily active users 200 thousand impressions that's about a 1 impression for daily active user ratio that's low we like to see a higher number than that but these guys are just getting into it and they were being pretty conservative so roughly 1 AD per user per day but let's look at that map a little bit deeper because that's not actually correct we're going to look at which players specifically we're generating those 200,000 impressions and the ratio is not at all one impression for daily active user because only 14% of the users that were in that game we're generating those impressions so that means that only 28 thousand players we're seeing and engaging with rewarded videos so that's 7.1 impressions per rewarded video viewer as opposed to one to one so you're seeing a lot fewer players seeing a lot more ads it's obviously far too few we want to expose as many players as we can to reward a

video especially those that we know are not going to make in-app purchases which is roughly 98 percent of your users to try and generate revenue and also generate engagement that comes from rewarded video so we got three steps that we want to go through here the first one is make rewarded video engagement a top KPI and what I mean by rewarded video engagement or the rewarded video engagement rate specifically is the percentage of total players viewing rewarded video so in this case it was 14% of all your players uh viewing rewarded video far too low you can get that number up quite a bit by doing a number of simple things so first of all you got to measure that rate so you have to know how many unique users are actually seeing rewarded video as a baseline case in this case 14% was very low we wanted to work on that there's only a handful of reasons that that's the case one is conservative initial ad rules so in other words especially when a lot of publishers are getting into ads for the first time they're pretty conservative around who

they expose rewarded video to they don't want to cannibalize a paying player they don't want to drive players out of the app so they tend to be very conservative when they step into showing ads look at those rules I mean we come across publishers all the time they want to wait till someone's been playing the game for three months wait till they hit level 50 etc etc all of the time those rules are too conservative and they leave a lot of users out of the loop in terms of being able to engage with rewarded video and ultimately they leave a lot of money on the table so first of all pretty restrictive rules of all they buried the rewarded video opportunities so sometimes they still see publishers putting it in the in-app purchase section of the app that if you want to earn a few virtual coins that you can watch a video burying rewarded video is obviously it obviously makes it a lot harder for players to discover and you end up with a very much lower engagement rate another thing that that we're working on more and more with publishers is making the rewards themselves more engaging so

as an example you might want to offer a reward of say gold or the standard currency early on in a game like cash yet you know 500 gold for watching a video well gold might be a pretty constrained resource early on in the game but as players get more advanced and level up they might end up with a lot of gold but end up with a different currency that's constrained so point is make sure that the reward you're offering is contextual to the user and valuable to the user because if you offer gold and gold is in abundance later in the game you're going to see a huge tail in terms of players engaging with rewarded video and we spend a lot of time with our platform helping our publishers segment users by their behavior and being able to offer them dynamic rewards or rewards that are different and contextual to where they are in the game and what currency they might need in the game so make it valuable to them and finally you know we're big fans of course of interstitial related video so instead of making a user hunt around and find where they can get virtual currency they're a button or

somewhere buried in the store make it interstitial so player hits a certain point in the game the odd opt-in comes up automatically hey you did this in the game do you want to see a reward video for this much currency don't make the players figure out how to get it make it part of the core game loop and and you'll see a lot higher engagement finally best practices and any ads tested creative and make sure that people are responding to the offer that you're making to them first and foremost just remove as many restrictions as you can we did that we drove at engagement up to about 40% 40% sabab rige we see across all the games in our platform it's not stellar but I think that's a reasonable benchmark to say 40% of your players are actually engaging with rewarded so what that does is it changes the base number we had 28,000 daily active users delivering on AB reviewing on average 7.1 impressions at a $10 CPM and making you $2,000 a day by improving your

rewarded video engagement rate we increase that number to 40% which is 80,000 diu we also know that these players are also going to behave very similarly to those initial 28,000 players in fact they're probably going to see more impressions and they're probably going to perform better from a CPM perspective for reasons I'll explain in a few minutes but for the base case let's assume they behave like the other 28,000 and they watch the same videos and they earn the same CPM so that just increased your CPM or your daily revenue from two thousand and fifty six hundred and eighty dollars a day let's dig a little bit deeper into that audience that are actually seeing rewarded video and look at what they're doing I mentioned earlier we have a 30 impression cap so once people see 30 rewarded video impressions a day they can't see anymore they aren't shown any more ads but how is that really affecting the user experience the engagement of those users and ultimately the revenue that you're generating

here's the actual chart of the number of players in the system as a percentage and the percentage of total impressions so you can see the giant line over to the right basically says that that's where we hit the cap and you can see that a massive number of users and the percentage of ads were generated at that cap in fact it to make it a little simpler than seeing the chart 30 percent of all the video rewarded video Watchers hit the cap and 50 percent of all those impressions we deliver the 568 thousand occurred at that cap so that's a really hard wall we're putting up for players that want to watch a lot of ads that aren't paying that are looking to generate virtual currency so the second step of our three-step plan is to remove or measure actually measure these caps and you can do that by a be testing two separate groups and say these guys are getting the cap these users are not getting a cap let's let it run for

a week or two look at those downstream metrics and make sure that we're not cannibalizing or having any negative metrics that we want to deal with I can almost guarantee you that you will not see negative metrics and in fact the uncap users will play more and they will generate a lot more ad impressions but pretty simple if you're using a reasonable analytics platform to segment those users and be able to evaluate whether you can or can't remove the cap by and large you can and should remove that ad cabin so we just extrapolated this is now what the distribution of views look like in terms of number of people watching ads and you can see that we it resembles more of a power curve we see that natural decay not this big wall and we see a lot more impressions being generated by a smaller number of people and the result of this particular in this particular case was 25% more impressions were delivered so that's another 142 thousand impressions delivered per day by that same audience

of players by removing the cap but unfortunately these impressions aren't making publishers any money and why well that's because these campaigns are the vast majority of them are performance campaigns and so I'm just going to give a quick rundown of how performance ads work just I'm sure a lot of you are familiar with it but I'm just going to use this simple example to illustrate the point and how we got to attend all or CPM and why those users are not making any money so I've used a game of war our pals machine zone obviously a very big performance advertiser these metrics are made up they're not real machine zone install or conversion metrics I just used them to illustrate the point so if a thousand users see a rewarded video approximately five percent that can go up obviously or very greatly five percent click-through and go to the App Store so you end up with 50 users those 50 users that go to the App Store four of them will ultimately download and launch and play that game companies like machines own pay on a per

install basis that's what a performance campaign is they only pay when they get an install so they track that campaign all the way back to the device and to the advertiser etc and they can see that those thousand views generated for installs that they paid two dollars and 50 cents or so they paid $10 we take the $10 we make balance that against the thousand videos that we saw and we end up with a $10 eCPM that's why it's called an e CPM because it's an effective CPM it's not an actual CPM like you would typically see in advertising so that's how we get to our $10 from a performance campaign the problem is when you have users that are watching 30 40 50 rewarded videos a day they are not clicking and they are of course not installing those games they are actually just watching those videos to earn currency when that happens we call those players for lack of a more affectionate term grinders grinders that are just watching video to get currency are never clicking they're never installing and all of those impressions mathematically are making you zero dollars so as we were

discussing 50% of the impressions generated or generated by these grinders that we're watching at least 30 videos a day all the impressions delivered no installs and therefore no revenue so we're going to do a little bit more math here and take those 568 thousand impressions we're going to split them 5050 two hundred eighty four thousand of those came from the grinders 284 came from people who watch less than 30 videos a day the 284,000 that were non grinders generated that amount of revenue fifty six hundred eighty dollars a day the other 284 thousand impressions delivered by grinders plus the 25% more impressions we got by removing the cap also came from those grinders which means there are four hundred and twenty six thousand impressions that generated no revenue for us on that day unfortunately there's a solution and this is where we get to a very emerging category of ads merging a little slower

than we would hope but it is now coming on strong brand campaigns and these are true CPM campaigns so that means that advertisers brand advertisers just want to show a video they're not trying to install an app or have any sort of performance criteria associated with that app with that ad other than they just want someone to watch the video so we've used coca-cola brand advertisers obviously are just trying to show you that video as a in the similar format that you see in a commercial they just want brand exposure and awareness they don't pay as much today as performance campaigns because the demand isn't there yet and there aren't enough advertisers and things like that but conservatively we're estimating that brand campaigns across the network can generate a CPM of about five dollars that's only about half of what you would get from a performance campaign but they pay that five dollars just by watching the video so whether you're a regular player that's going to install or whether you're what we call a grinder is watching tons of ads a day they still pay that same CPM rate so step three of the three-step plan is to

deliver CPM campaigns only to users that are unlikely to convert on CPI campaigns as we've called them grinders so we're going to segment our audience into two distinct groups to do that we have to track user level out engagement to determine how many views which people are seeing that many ads and then we have to come up with that threshold to illustrate the point here we've called a grinder anyone who watched swatches 30 or more rewarded videos a day that that's just our definition there are way more advanced ways to determine that we can track clicks and things like that but let's just for the sake of illustrating the map say that anyone who have used 30 or more rewarded videos a day it's not installing anybody else's app they're just doing it for the currency the reality is there's probably a lot more what we know there's a lot more wasted impressions there and we can do a lot more advanced things to find them but for those grading the point here let's assume that those players 30 ads plus a day are not installing and we're going to give them CPM campaigns so the result is that those four hundred

and twenty six thousand impressions you were getting absolutely zero for you're now getting a CPM of five dollars so we had a case where eighty thousand daily active users we're generating you fifty six hundred and eighty dollars off of five hundred and sixty eight thousand impressions we know a lot of those impressions were worthless for performance so this math because we split them is no longer relevant instead we're going to look at how the math actually works once we've segments in our audience so we take those people that watch less than 30 rewarded videos and we keep showing them performance campaign so they're generating that same fifty six hundred and eighty dollars but we're going to show branded ads to the grinders which is four hundred and twenty six thousand Preston's at $5 which generates another 20 130 dollars in incremental revenue so now we've increased our revenue to seventy eight hundred and ten dollars a day so to recap three steps start with two thousand dollars a day you drive higher rewarded video add engagement by loosening up the ad restrictions and

measuring the impact and making sure that more of your audience 40% or more ideally are watching rewarded video second take a look at things like cap walls and evaluate them and remove them in most cases the vast majority of cases every games a little bit different but I would say by and large there's a 99% chance that an ad cap is hurting you and not helping you remove that to drive more impressions and finally segment your users into performance campaign and branded campaign viewers and show branded campaigns to users that are not converting on CPI campaigns in other words you're grinders and the end result is you end up going from two thousand dollars a day to seventy eight hundred and ten dollars a day off the exact same audience you've improved your engagement metrics because people are staying and playing and learning more currency that they're that they're spending in the game and that results in another fifty eight hundred and ten dollars a day in revenue or more than two million dollars a year in incremental revenue by doing

those three things that's what we do at up site shameless plug our director of at operations handsome Barry D'Elia he came up with a lot of these wacky ideas and we have a full team that's responsible for working with our partners to drive this kind of experience and make sure that our publishers are not leaving any money on the table so that's it for me thank you guys I mean [Applause] [Music]