If you have been following the SOOMLA blog, attending mobile game conferences or keeping up with the latest mobile monetization trends in some other ways you should already know the following important fact. Improving Opt-in to rewarded videos usually results in an increase of the same proportion in your total ad revenue. This is why many companies that use rewarded videos have been focusing on the opt-in parameter and have been trying to optimize it.

While getting the basic opt-in ratio is easy, there are a few advanced methods for finding hidden opportunities around opt-in rates.

1 – Look at daily opt-in vs. monthly opt-in

Typically, app companies focus on the monthly opt-in – this is the ratio that is normally available by platforms such as Ironsource mediation and what most will allow you to analyze if you send them the impression events. The monthly opt-in, however, only tells part of the story and in many cases we have seen that the daily opt-in can be significantly lower. What that means is that there are users who opt-in to the videos some of the days while not watching videos on other days. Fixing this can usually yield 20-25% more in ad revenue and the way to do it is by taking a close look at your incentives. Will the users need the incentive on a daily basis? If not, try to figure out an incentive that the users will need more regularly.

Definitions
Monthly opt-in – the number of unique users who watched at least one video in a given month out of your total MAU.
Daily opt-in – the number of unique users who watched at least one video in a given day out of your total DAU. The daily opt-in has to be averaged across multiple days to smooth out the fluctuations.

2 – Analyze opt-in for cohorts

Cohort analysis is hardly a new trick for marketers but when it comes to monetization managers it actually is. Comparing the opt-in rate for new users vs. existing users can lead to some pretty interesting insights based on our experience. This might requires some help from your BI team (or simply using SOOMLA’s dashboard) but the hidden opportunity should justify the effort as we have seen up to 2x differences between the two segments. If opt-in is high for new users and declining for long-term users it could be a sign that your incentives are not meaningful enough for your users. In other words users are willing to watch videos but they soon realize that what they are getting in return doesn’t get them very far so they stop. In other situations, the opt-in for new users is low. This could indicate an awareness and training problem. Making your users aware of the option to watch videos early on can fix the problem.

3 – Differentiate users from different traffic sources

One of the interesting patterns we have seen is that users from different traffic sources behave differently when it comes to opt-in ratio. Users who came from paid channels and specifically from video ads often present a higher opt-in ratio compared to organic users. To improve the opt-in ratio for organic users, consider adding some more guidance to highlight the opportunity of watching videos for in-game rewards.

4 – Treat your ad whales to nice Incentives

In recent research we showed that the top 20% of the users contribute 80% of the ad revenue. These so called Ad Whales are the most important segment from an ad revenue perspective. You should focus a lot of your attention to make sure the opt-in rate for this group is as high as it can be. These users typically contribute more than $0.99 and sometimes up to $100. This means that they are as good as payers and you can offer them in-game items that are normally reserved for an actual purchase. However, since you want users of this group to watch a video daily it’s better not to offer them a perpetual item. Some examples of incentives you can give for ad whales:

• A tank that is normally worth $100 – watch a video to use for a single day
• Shortening a waiting time that normally costs up to $1
• 10x coin boost for a short period instead of 2x

Identifying the ad whales is possible by attributing the ad revenue accurately to the user level. The only way to do this accurately today is with SOOMLA Traceback.

We’ve put out a series of posts on the wide topic of Opt-In rates and the importance of them. Feel free to check them out:

• Increase Rewarded Video Revenue by Mastering Opt-In Ratio
• 10 Ways to Improve Opt-In Ratio for Rewarded Video
• Measuring and Improving Opt-In Ratio with SOOMLA TRACEBACK

If you downloaded a match 3 game recently you might have noticed that the first few levels are guiding you how to play the game. I would make a guess that if you are reading this blog you have already played several match 3 games already and the tutorial is not really relevant for you. The question is how frequent is this situation and if there is a better why.

Match 3 – a popular genre

Match 3 games are a popular genre on mobile – no doubt about that. A Google site search on the play store returns over 1M results for the search “Site:play.google.com “match-3”.

It’s probably the genre with the most number of apps out of the narrow genres (as opposed to a wide genre such as “Strategy”).

If we look at the number of players: Candy crush alone has over 2.7 billion downloads – this is close to the number of app capable devices out there and there are many more apps who are not that far behind.

It’s not a new genre either. Match 3 games has been around for at least 15 years. They have been with us through web games, Facebook games and mobile games.

A new user in a match 3 games is not new to the genre

What all these stats mean is that if your company publishes a match 3 game it’s likely that the majority of the new users you are getting is already familiar with the genre. Your UA teams are targeting users who liked other match 3 games on Facebook, Google is targeting your ads to users who searched match 3 in the past and other networks are trying to achieve the same result to send you relevant users. Users who know how to play match 3.

The tutorial is redundant for experienced users

Showing a tutorial to a user who never played match 3 could be the difference between him staying or leaving. However, the same tutorial for an experienced user is not effective. In fact, it might have the opposite effect – causing the user to leave as he is not challenged enough. Consider the tutorial in the image shown on the right side – it basically says “match 3 items” and might evoke the reaction “Well Duh!! It’s a match 3 game”.

Detecting experienced users automatically

Game publishers who want to offer an adaptive tutorial experience face a new challenge – how to detect which users are experienced match 3 players vs. not. Here are a few ideas how to detect this:

• Acquisition channel – normally your UA efforts will be targeted at match 3 fans so you can just treat all paid traffic as experienced players. Alternatively, separate campaigns that are directly targeted to match 3 fans vs. broader campaigns. Using deeplinking you can invoke different flows inside the app. 
• For the android version of your app, you can potentially check what other apps are installed on the device to determine if the user is already familiar with the genre. While Google will not allow you to send the app list to your server, checking locally and adapting the game experience is in the benefit of the app publisher as well as the user.
• Prompting the user and asking him if he knows the genre is another way to go. If you think asking the user questions is annoying you should think again. Going through 7 levels of learning the game is far more annoying.

As SOOMLA is the 1st company to focus on monetization measurement for mobile apps it only make sense that we will take the lead on predicting some of the trends that will control app monetization in 2018. These predictions are based on our data as well as on observing the market trends in 2017. However, predicting the future is a tricky business so take these with a grain of salt. Here we go – counting 7.

1 – Let the ad whale hunting begin

In 2017 SOOMLA exposed the existence of ad whales – a group of users who contribute over 80% of the app ad revenue and can sometimes make $100 for the app publisher by watching and interacting with ads. In 2018 more and more app companies will invest resources into understanding who the ad whales are, how they behave, what are the best channels to bring more of them and how to adapt the app for this segment. Basically, the same practices app companies applied for the top spenders will be used for the users who generate the big advertising dollars.

2 – Publishers will seek tighter control on ‘rich’ interstitial ad content

2017 introduced a lot of innovation around ad-formats that can be delivered through interstitial containers: Playable ads, interactive videos, dynamic end cards and what not. Publishers who integrated interstitial ads expecting a short ad break in the app flow ended up with an experience they didn’t sign up for. The fact that a longer ad experience with an invisible ‘x’ button has a toll on retention intuitive but SOOMLA also validated that with data and will publish a report about it in Q1/18. In 2017 some publishers started pushing back on these formats and we expect more publishers will want to control these ad experiences in 2018.

3 – More publishers that are also advertisers

In 2016 there were very few ad driven app companies that could afford paid UA campaigns. In 2017 this number grew and in 2018 it will grow even more. Following the footsteps of SOOMLA, more providers are offering tools that give visibility into Ad LTV. In turn, more publishers are aware of where they stand and what CPI levels they can bid. See the post about the steady increase of CPIs and how they are here to stay.

4 – Header bidding will start but adoption will be slower than expected

Header bidding was discussed in many conferences in 2017. The idea is simple and highly beneficial to publishers and some ad providers have launched earlier versions of this model. In 2018, some publishers will test out this model but it will not go into mass adoption just yet. There are too many loose ends at the moment and no sufficient coverage from ad providers. Furthermore, it appears that the some of the leading players are happy to receive bids from others but no so happy to provide the bid out. FB, Google, Mopub, Appodeal and Ironsource are each trying to become the company who will run the auction so they refuse to give a bid out. This means each that each one of them insists on exclusivity which will be a big turnoff for publishers.

5 – Better control over ad experience and creative

Publishers needs ways to control the ad experience as part of the overall app experience. In 2017, SOOMLA and SafeDK started providing solutions in this area. We expect more solutions will become available, more publishers will start using these and ad providers will also start adding more functionality to control ad experience.

6 – More apps will advertise competitors in 2018

Advertising competitors was a big no-no for many app publishers who were concerned their users will churn away and move to the competitor app. In 2018 there are already tools that allows monitoring the eCPM and churn caused by specific advertiser. This means app publishers will be able to apply a data driven approach to this question that was decided with gut instincts until recently. Based on the data we have seen – more publishers will feel comfortable with advertising competitors as a result.

7 – Ads will surpass IAP for mobile game monetization

2017 ended up with a tie between the different monetization models for games. Some studies claimed IAP revenue was still bigger while others showed ad revenue as the winning monetization model. In 2018, there will be no question any more and the clear monetization winner will be ad revenue. Part of the reason for that is the emergence of data tools to measure ad monetization. This makes more publishers feel comfortable with building games that relay heavily on ad revenue.

That’s it – 7 predictions for the new year. Write them down and check if we were right in 12 months.

Rewarded videos have been proving to be a highly effective format that balances retention with monetization and typically turn eCPMs of over $20 in the US. However, unlike other ad formats, they require the game designer to build incentives that can be handed to the users when he watches videos. Check out our previous post that talks about mastering the Opt-In ratio to boost rewarded video ad revenue. The general rules of thumb on where to implement rewarded videos are:

• Designing the incentives from the beginning is easier than adding them after
• The more evolved the meta game is, the more opportunities for videos there are

Here you can find the most popular incentives to entice users to watch videos:

1. Lives or “save me” option

This is a familiar incentive that allows the player to cheat death. It it widely used in Match-3 games where the meta game typically dictates that a player may only fail 5 times before he runs out of lives at which point he typically has to wait for his lives to replenish. Another version of this incentives appears in action/arcade games where a violent death of the character typically ends the player’s session and an option to keep the session going makes a strong incentive for the user to watch videos.

2. Time related incentives

Warping the game time can be a compelling incentive for the player. In some games, the player is only given a limited time to complete a mission or a screen and is offered to watch a video to gain more time. In other situations, the player wishes to avoid a long wait and is willing to trade that wait for watching a video. A perfect example is upgrading a piece of equipment which takes 2 hours, but if you watch a video ad, it upgrades immediately.

3. In game currency

This one is simple and quite obvious. A bag of gold in exchange for watching a video is one of the oldest offers out there. It’s not the most effective incentive but it’s widely applicable and typically doesn’t require any special hooks to be put in the game for it.

4. Earnings doubler

The coin doubler is known as a paid item that players can buy. However, a limited time doubler may be offered as an incentive to players who are willing to spend 30 seconds watching a video ad. This type of incentive is popular in idle clicker games and runner games among other genres. There are 2 variations of this incentive:

• Pre-session – allowing the user to start a session knowing his earnings will be doubled
• Post-session – popping the question to the user in the session summary screen

5. Re-dealing of a randomly assigned element

In many cases a player get dealt a hand of cards, in other games he opens a pack of collectables and sometimes it’s quests, missions or even songs that the game randomly selects for the player. If a player doesn’t like his options, he can change them in exchange for watching a video.

6. Renting items

Some items in the game can be priced very high and not many users can afford them. Giving them away for a video watch can reduced the perceived value of such items. The compromise is to rent such items for a limited period of time. If an item costs $50 and you expect the users to use it for 30 days renting it for 15 minutes for a video view is maintaining the balance and unlikely to hurt your conversion to payers.

7. The daily spin

Many games offer a daily spin or surprise box as part of their meta game. It’s a great way to make players feel welcomed and keep them in the game. This daily spin often ends with a near miss experience and users are likely to watch a video ad if one is offered in return for an extra spin.

Some of our customers started asking us about GDPR, so we created a GDPR Compliance FAQ page to answer most questions. To those of you who haven’t heard about it, GDPR is the new European privacy regulation that will take affect on 25th of May 2018. The new set of rules is causing many companies to lose sleep also due to the recent Disney lawsuit and the lawsuit against Kiloo, maker of Subway Surfer. The first suit includes not just the company itself but also tech providers such as Kochava, Upsight and Unity.

This means that while the app companies are on the front, the technology companies behind them should also learn GDPR carefully. In the rest of the post, I’ll try to describe some of the key differences and explain what actions companies should consider. So lets jump into what’s new with the GDRP:

The price tag difference

One of the new aspects of GDPR is that it names a price for non-compliance – fines can reach up to €20M Or 4% of annual gross revenues – the greatest of the two. What this means for app companies is that they have a strong business case to invest money and effort in complying. For tech companies, it means that the level of liability will be pushed up. If tech companies were able to get a way with capping their liability at a $0.5M or $1M dollar, that will no longer be acceptable by the app companies.

US companies also on the hook

Another key difference is that GDPR makes it clear that as long as companies have users in EU, the rules apply to them regardless of their location. For US companies, this is a major difference as privacy rules in the US are less restrictive.

Everything is personal

One thing that the GDPR makes very clear is that all device identifiers including IDFA (Apple devices’ ID for advertising), GAID (Google’s advertising ID) and IP address are now considered personal data and any data stored with it in the same record should also be considered personal. This have been a gray area a few years ago but was getting less and less gray in recent years. With GDPR there is zero doubt about this. For app companies and, advertising companies and analytics companies this means that all data becomes personal and should be treated as such.

Encrypting and protecting and documenting data transactions

Another requirement that GDPR makes more clear is the need to encrypt and protect personal data as well as document any transaction in which encryption was not possible. This is not a new requirement but since all data is considered personal now it becomes a requirement for each company and each piece of data. Here is an example of one process that is likely to change for App companies and has already affected some of the tech providers. In the past, companies used Facebook’s highly effective lookalike modeling service by creating custom audiences based on divide identifiers. The practice of exporting a CSV file from you analytics or attribution platform, storing it in your personal computer and uploading it to Facebook is now considered a non-encrypted transaction that has to be documented. Not many app companies will want to cumbersome their process with the documentation requirement and so some of the attribution companies have responded with audience builder tools that make this transaction encrypted.

Is your data coming to US for business or pleasure

Another area that app companies and tech companies serving them should be aware of is the transfer of information outside the EU and specifically to US. This has been a key topic for previous legislation but the requirements became stricter with GDPR. This topic is known as cross border transfer. In a nut shell, EU knows that US is more liberal when it comes to privacy and specifically in the Federal’s government ability to force companies hand in providing private data. One example of the FBI power over companies was last year when Apple confronted the FBI and refused to help them crack an iPhone. While Apple stood up to the FBI, very few companies will risk disobeying a court order.

To adapt to the new regulations, companies can no longer rely on gaining the user consent for transferring their data. This practice is required but not sufficient anymore. Instead companies should do one of the following:

• Keep data about EU users in Europe and comply all tech providers to do so
• Make sure all providers are part of the Privacy Shield initiative
• Execute model clauses to document each data export from EU to US

Keeping data in EU

This may sound easier than it is. If you are an app company, you are probably using at least a dozen services to help you monetize users, analyze them and improve your app. Most likely, you also have homegrown analytics tech that reads and writes data to a database stored by a cloud provider. Keeping the data in EU means you need to go to your cloud provider and each one of the other tech provider and make sure they also keep the data in EU. In turn, the tech providers will have to go to their tech providers and cloud providers and do the same. While the major cloud providers: AWS, Google and Azure have data centers in Europe it’s unlikely to ensure 100% of the data staying in EU given the number of providers involved especially when the app is serving ads.

Privacy Shield

This is essentially a certification that companies can get if they do store their data in the US. Being listed in the Privacy Shield list of certified companies is an alternative requirement to keeping data in EU. It means that tech providers who don’t keep EU user data in EU can get a Privacy Shield certification and help the app companies comply with GDPR. In the list below, you can find popular SDK providers that already obtained Privacy Shield certifications and the ones who didn’t. The extensive list can be found here – https://www.privacyshield.gov/list . Note that:

• Providers that store data in EU don’t need the Privacy Shield (e.g. Adjust)
• Providers can give an EU model clause document to app companies as an alternative to Privacy Shield and still comply with GDPR.

SOOMLA is already in compliance with the regulations and has started a process to obtain a Privacy Shield certification. Ask us about the current status by emailing – privacy@soomla.com.

EU Model Clause

As mentioned before, providers who don’t keep their data in EU and don’t have a privacy shield certification can still help app companies comply with GDPR by providing an EU model clause document explaining exactly how and what personal data flows from EU to US.

The right to be forgotten

Another key requirement by GDPR is that every user has the right to be forgotten. This means that a user can request an app publisher to delete all his data including data about him that is stored by 3rd party providers.

What you should ask from your providers

While there are a few changes app developers might have to implement in how they handle their users’ data most of the work will probably be with ensuring their providers’ compliance and more specifically their advertising related ones. The decision to treat IDFA, GAID and IP addresses as personal data puts all the advertising industry in the spotlight as most of it was operating under the assumption that IDFA will not be considered personal data.

Here is quick compliance checklist for your providers.

• Do you protect and  encrypt any record that contains IDFA or IP address?
• Can lists of IDFAs or IP addresses be exported? Do you send such lists over email?
• Do you keep the data in EU or US? If in US – are certified under privacy shield? Can you provide model clause explaining all data transactions?
• Forgetting users – Make sure you know all instances of the user record (log, backup, main DB, …) and what’s the mechanism to delete.

Providers Not Certified with Privacy Shield [updated Nov-2017]

• Appodeal
• Chartboost
• Ironsource
• Fyber
• Adjust (But already has a stricter certification)
• Heyzap
• Lifestreet
• Media Brix
• AOL / Millenial
• Tapjoy
• Vungle
• Upsight/Fuse
• Unity / Unity Analytics / Unity ads

Providers Certified with Privacy Shield

• Google – including Cloud, Admob, Analytics and Firebase services
• Amazon – including AWS and Amazon Ads
• Microsoft – including Azure
• Applovin
• Adcolony
• Appsflyer
• AppAnnie
• Mixpanel
• Facebook
• Kochava
• Amplitude
• TUNE
• Mopub

App Publishers who monetize with ads often face the need to compare between ad-networks. Which one offers stronger monetization? Is the network declining in strength? Who should i put first in the waterfall? The common practice today is to look at the average eCPM but actually looking at the 1st impression eCPM is a much better approach. Here are 3 reasons for that.

Networks put their best campaign first

Each ad network has internal optimizations mechanisms in place. Some have algorithmic approach that try to predict the eCPM of each potential ad given who is the user and all the data they have about him. Others have more simplistic priority lists. Either way, when the network sees the user for the 1st time in a given day, it will try to put the best ad for that user. In later impressions, they have to circulate in other ads, their 2nd best, 3rd best and so on.

Average eCPM is a self fullfilling prophesy

Average eCPM on the other hand is influenced by many parameters other then the network’s stregth. In situations where the average eCPM is used to determine the priority between the networks it acts as a self fullfilling prophesy. To understand this, let’s look at the two ends of the priority list:
The Network with First Priority – This network gets more 1st impressions than any other network as long as it has fill for them. This drives the average up. At the same time, the network also wants to stay at the top and knowing that the publisher is looking at the average eCPM it is likely to set a price floor that will eliminate the low eCPM campaigns. This will also drive the average eCPM up.
The Network with the Low Priority – This network is getting less 1st impressions so their average will be lower. Even if the network landed a major campaign it will not get a lot of exposure and will not be able to drive the eCPM up. At the same time, the low priority network can’t shut down the low eCPM campaigns as that will completely choke the delivery for their advertisers and will cause a new bag of issues for the network.

Changes in 1st impression eCPMs are clear triggers for action

Tracking different parameters is a good practice but tracking becomes much more powerful when it’s connected to actions. When you track the average eCPM and you see a drop in that paremeter for one of the ad-networks there could be a few potential explanations. For example, if that network is getting a high percentage of later impressions it would bring down the average. The 1st impression eCPM is less influenced by how you are using the demand source and is a better indicator of the quality of the demand. A drop in the 1st impression eCPM can be caused by the ad-network losing an important advertiser or by them changing the rev-share on their end. Either way, it’s a good reason to look for new partners to take the lead.

Tracking 1st impression eCPM – Easier than ever

The reason why more company focus on average eCPM rather than 1st impression eCPM is that this is the information the ad-networks are making available on their dashboards. Publishers that use SOOMLA, however, have easy access to reports about the 1st impression eCPM over time and the 1st impression eCPM of every single campaign by each ad-network in addition to the average eCPM.

Many companies ask themselves these days if they should be hiring a Monetization Manager now or wait until it’s volume is larger. In this post we will try to provide a simple framework for thinking about this question.

Ads first games vs. IAP first games

There are two types of companies to consider for this question. Before you continue, you should ask yourself which type of company are you. The framework for evaluating the merits of hiring a monetization manager differs a bit between the two types of companies. Here is the profile for each one:

Ad first games – These are typically smaller companies. If you are an ad supported company and still debating the monetization manager question it’s unlikely that you have more than 15 employees. These companies tend to have a mix of at least 3 ad formats from this list: banners, native, interstitials, video and rewarded video.

IAP first games – These are typically more established companies who already do well with IAP and treat ads as a secondary channel. The ad formats in use here are mostly rewarded videos and sometimes offer walls.

The basic formula

There are 2 conditions to be met before you hire a monetization manager:

• The ROI condition
• The focus condition

The ROI condition

[monthly ad revenue] x [improvement opportunity ratio] x [risk factor] > [monetization manager full cost]

Where:

• Monthly ad revenue – how much your app is making every month from advertising
• Improvement opportunity ratio – Estimation of how much you can improve
• Risk factor – the chance of that improvement actually happening
• Monetization manager full cost – Salary + social benefits + taxes + direct overhead increase + cost of tech tools + cost of projects he will drive

The focus condition

The focus condition is looking at the same formula but instead of justifying the direct cost, you are estimating the opporunity cost. The focus condition is more relevant if you are projecting that the monetization manager will be driving many requirements to R&D and BI departments. We will see how to evaluate how much effort the monetization manager will require in the paragraphs below.
The way to think of opportunity cost is usually top down. Let’s say that the goal of the company is to double in revenue within 12 months. This means that each quarter you are looking to get 20% growth. Most companies can’t contain more than 2 focuses each quarter and some say 1 is enough. This means that if the monetization manager and all the tasks associated with him will not generate 10% increase it’s not meeting the focus condition. The formula will look as follows:

[improvement opportunity ratio] x [risk factor] > [Required quarterly improvement] / [Quarterly initiatives count allowed]

Estimating the improvement ratio

For IAP first games

• Improving opt-in ratio for rewarded videos – high product and R&D effort – can double or triple ad revenue when combined with A/B testing.
• Adding more demand partners – medium product and R&D effort – the improvement in ad revenue can be up to 50% depending on current status (see full explanation below)
• Applying CPM price floors and cutting fixed CPM deals – no R&D effort – up to 15% improvement
• Blocking low eCPM advertisers and optimizing volume for high eCPM ones – no R&D effort – up to 15% improvement
• Setting different ad strategies for different segments – low R&D effort – up to 30% improvement
• Acquiring users who respond better to ads – no R&D effort – up to 50% improvement

For Ads first games

• Optimizing the frequency and mix of ad-formats – medium R&D effort – can improve ad revenue up to 50%
• Adding more demand partners – medium R&D effort unless done as S2S – the improvement in ad revenue can be up to 50% depending on current status (see full explanation below)
• Applying CPM price floors and cutting fixed CPM deals – no R&D effort – up to 25% improvement
• Blocking low eCPM advertisers and optimizing volume for high eCPM ones – no R&D effort – up to 15% improvement
• Setting different ad strategies for different segments – mid R&D effort – up to 30% improvement
• Acquiring users who respond better to ads – no R&D effort – up to 50% improvement

How much your ad revenue can improve by adding demand partners

The improvement ratio per ad-format is driven by how strong your demand and fill rates are currently. We included a basic formula that we found helpful but you should do a better job assessing this by looking at specific countries and diversified demand. We also recommend Jonathan Raveh’s post on this subject. Here is a simple formula to start with:

(2x[number of ad-networks serving banners]+1)x[banners revenue ratio from total]/2x[number of ad-networks serving banners]-1

Estimating the cost of the monetization manager

$8K/month or $96K per year is a nice salary for a monetization manager in US. The taxes and benefits in US can come to 25% to 40% on top of the salary. Office space and immediate overhead per employee can be around $500 based on WeWork rates. In addition, we should add the average license cost of SOOMLA ($3,000) since having a monetization manager and not giving him the right tools to optimize would be moot. The total comes to $13,500 – $15,000.

Estimating the risk

The risk ratio is slightly harder to estimate. You should think of all the things that can go wrong and try to assign probabilities. Here are some items to consider:

• Bad hiring can set you back
• If you can’t afford a SOOMLA license your risk will be higher
  • The monetization manager will not be able to a/b test the ad revenue so optimizations might have a negative impact
  • His ability to set the right price floors will be limited
  • He will not be able to analyze and optimize on a campaign level
  • Segmentation will not be possible for him
  • The users that are being acquired by the UA team will not be a good fit for ads
• IAP first apps monetize mostly with rewarded video where negotiating eCPM price floors with ad-networks is only possible for high volume apps.

Example – finding the ad revenue threshold for hiring

Let’s look at one example of using the formula. We can estimate that the total opportunity to improve is 60%, the risk factor is 50% and the total cost of the monetization manager will be $15,000.

[monthly ad revenue] x 60% x 50% > $15,000

To satisfy this condition we need an ad revenue of at least $50K / month or $600K annually. The numbers we choose are reasonable so if you have this level of ad revenue and you are not hiring a monetization manager you are probably leaving money on the table. Of course, if you have $1M/month from IAP and only $50K in ad revenue, you might have bigger fish to fry first. This is where the focus condition comes in to play. Make sure you evaluate both before you make the decision.

When integrating ads, one of the biggest concerns is that users might churn away. There is an obvious trade off between the need to give the users a great experience and the need to turn revenue. Not all ads are created equally when it comes to their impact on user retention and it’s important to measure the impact of different ad types and monitor what ad experiences your users are getting from your ad partners. Below is a list of ad experiences to watch for:

1. Crashes and freezes can impact mobile app retention

Ads are served by the ad networks who tend to require an SDK integration. Each SDK increases the complexity of the app and might conflict with other SDKs, in turn cause your app to crashes for your users. This happens especially in edge conditions such as old Android versions or uncommon devices. While crashes are typically reported through your crash analytics provider, there is another type of error that is trickier to track. In some cases, the SDK of the ad network will try to show an ad to the user but will end up freezing the device. This type of error is typically not detected and is harder to monitor but it could have the same negative impact. Both of these errors might cause users to churn away and reduce the overall app usage experience.

2. Close buttons that are hard to find frustrate users

3. Lack of ad diversity will bore your users

4. Poorly targeted ads may get your users annoyed

5. Your users don’t want to wait for a slow loading ad

6. Deceiving ad creatives are hard to tell from your app buttons

7. Inconsistent ad skipping experience and long duration ads

Video ads are becoming an increasingly important monetization format. Even the biggest app companies are utilizing video ads as part of their monetization strategy and specifically, mobile gaming companies have widely adopted the rewarded video ad format that provides a positive experience for the user and is positively correlated with engagement and retention according to a few researches.

In this post you will find a list of the top 12 rewarded video ad providers divided into 4 categories:

• Video only networks
• Ad networks that moved strategically into rewarded video
• Video ad networks with a mediation platform
• Media giants who recently moved in

Video Only Networks

These ad networks are purely focused on monetization through video ads. They don’t offer any other ad format and some of them played a major role in educating the market on the benefits of rewarded video ads.

Vungle

Vungle are a key contributor in popularizing video ads among mobile app publishers. When they started out they were focusing on 15 second videos and were offering to produce the videos as part of the deal. Vungle is a private company and is backed by a long list of investors and raised $25M to date.

Adcolony

Adcolony is the first company to offer rewarded video ads for mobile apps and they are still one of the top providers in the field. They are 100% focused on video ads and are high on the list of any app publisher who wishes to monetize his app with video ads. Adcolony was acquired by Opera in 2014 for $350M but remained a seperate entity.

Unity Ads

Unity Ads came to life through the acquisition of Applifier by Unity. Since the acquisition, the video focused ad network experienced fast growth leveraging the dominance of the Unity game engine in the mobile space.

Receptiv (formerly Mediabrix)

Receptive who are also known as Mediabrix prior to their rebrand, have a unique offering compared to the last 3 companies mentioned. The company is based only on brand advertisers and has its head quarters in NY where they can be close to the media agencies. To the advertisers, they offer the opportunity to be associated with the winning moments of the user inside the game. To the publisher they offer diversified demand with high eCPM.

Ad networks who moved strategically into rewarded video

Applovin

Applovin was making waves in the ad-tech space last year by announcing it’s acquisition for $1.4B. The deal was experiencing some trouble and was not finalized as of today [July 2017]. Regardless of the acquisition, the company is operating as a seperate entity either way and is doing well financially. On the advertiser side, the company offers more control compared to other networks through their self-serve interface. On the publisher side they specialize in interstitials and video ads.

Chartboost

Chartboost started it’s way as a marketplace for direct deals and was one of the main contributors to the adoption of interstitials as a tool to promote games within other games. Chartboost came a bit late to the video ads space but were catching up quickly by leveraging the distribution of their SDK.

Tapjoy

Founded in 2007, Tapjoy made a significant transition from incentivized offers to video ads back in 2012. Currently they have been boasting impressive growth, grabbing larger chunks of the mobile advertising market share. From recent statistics, Tapjoy is estimated to be used in over 20,000 mobile apps.

Video ad networks who also provide mediation

Supersonic / IronSource

Supersonic became part of IronSource via the all Israeli acquisition valued at $250M. Together they are now considered the leader in mobile video mediation. In addition to the mediation service they also have their own video ad network which helps publishers top their fill rates.

Fyber

Fyber started as an offer wall provider by the name of SponsorPay but later on rebranded as Fyber and shifted more of it’s focus towards SSP and mediation with a strong emphasis on video ads. They acquired competing mediation service Heyzap to become a close second to fast growing IronSource / Supersonic platform.

Media giants who recently moved in to the video space

Admob / Google

Google needs no introduction and their mobile ad service Admob which became part of Google through the $750M acquisition in the early days of Smartphones is today the dominant way to monetize apps on Google Play. The giant rolled out rewarded video ads in March 2017. While they are showing later for the party we are sure that their size will allow them to gain momentum quickly.

Facebook Audience Network

Facebook dominates as a destination site for mobile ads but in recent years they have been evolving an ad network by the name of Facebook Audience Network and as of June 2017, FAN is also offering rewarded video ads.

Mopub / Twitter

Mopub was acquired by Twitter in 2013 for $350M (read more here). It kept it’s identity since and is one of the top 2 mediation platforms and and SSPs in mobile apps when it comes to banners interestitials and native ads. They showed up a bit late to the video ads space and launched video ads marketplace and mediation towards the end of 2015. Their stronger push in the video market only happened in 2017 however.

Header bidding created a big buzz in ad-tech spaces and the mobile app eco-system could not stay indifferent to it. There are quite a few problems slowing down the adoption of header bidding in mobile apps and it’s even possible that it’s the wrong model for mobile apps.

Header bidding – what it is

Header bidding works a bit like the role of SSP in the RTB model but different. In both SSP and Header Bidding – the publisher wants to get the best price for an ad impression that will be served to the customer. He runs an auction between the potential advertisers. Each advertiser submits a bid and the winner gets to serve the impression. This process is repeated for each impression.

There are a few differences however:

• In SSP the auction is managed on the server side and in header bidding it’s on the client side
• In SSP the winner pays the price of the highest losing bid (2nd price auction) while in header bidding the winner pays the full price
• Header bidding allows combining a few SSPs in the same web page or mobile app
• Direct deals can be treated according to their actual CPM and be added to the auction

Why was header bidding created in the first place

The HB and SSP models are so similar that one might wonder why header bidding was created in the first place. This is partly related to unfair behavior by some SSPs. Specifically, Google was mentioned in a few conversations I had about the subject. The most popular SSPs including Double Click by Google has their own horse in the race – for Google that horse is Ad-x. Any SSP that is running the auction but at the same time placing a bid has motivation for to bend the rules. Real time bidding might appear to be a transparent process in which bending the rules is harder, however, when there is a will there is a why. Specifically in Google’s case it was a feature called “enhanced dynamic allocation” that allows Ad-x to cherry pick inventory from auctions being run by Double Click (their SSP) by seeing the other bids first.

Header bidding in mobile apps

As you can guess from the name. Header bidding was created for web pages and “header” refers to the part of the html code that is loaded first. As of July 2017, none of the top 200 mobile apps has implemented header bidding according to our checks and most vendors who focus on mobile apps as opposed to mobile web don’t support pre-bidding at the moment.

3rd party vendors moving in but diminishing the benefit

Of course, the opportunity for in-app advertising is huge and the players are giants such as FB, Google and Twitter among others. With billions of dollars on the table, there are strong forces who try to push the mobile app eco-system towards header bidding. This can benefit DSPs who are interested in more direct access as well as the exchange providers. However, the adaptation of header bidding to mobile apps is not trivial and some of the offered solutions are “Header bidding in a box” where the auction goes back to the server side. This of course, diminishes the benefits of header bidding as the auction is outsourced to a party that may have bias.

Mobile app advertising is CPI driven

There is a bigger problem that is clouding the future of header bidding in mobile apps. It is not even certain that header bidding can be applied successfully? One might be surprised that not many mobile app companies are pushing for header bidding despite the trend that it created in the mobile web and desktop space. The situation in mobile app advertising is a bit different than that of web advertising. Specifically, mobile app monetization relies heavily on CPI campaigns. These are campaigns that pay only if the user installed the promoted app after he watched an ad. On the other hand, header bidding requires all the parties who are interested in placing the ad to come up with a bidding price upfront. This creates an adoption problem for header bidding. As of now, not many CPI networks are willing to commit to an upfront bid before they know what their payout is going to be. At the same time, mobile app advertisers got very comfortable with the CPI based model as it minimizes the risk. On top of that, for header bidding to work it’s not enough that one CPI network will send bids upfront. You need all of them to do it. This creates a critical mass problem and no one benefits from being the first one to move.

Someone has to take the risk

Going back to the dilema of advertisers that want to pay per install and publishers that wants to earn per impression. This is one of the oldest struggles in advertising:

• In CPI or CPA models – the publisher takes the risk and the advertiser enjoys guarnteed results
• In the CPM model – the advertiser takes the risk and the publisher enjoys guarnteed payout

CPC used to be the middle ground but click fraud killed it and the only ones that can afford to do it is Google due to size, brand and massive investment into fraud prevention.

If header bidding gains traction while advertiser continues to pay CPI, the risk will have to be taken by the ad-networks. For example, the ad-network might be bidding $5 CPM. Let’s say they serve 1,000 impressions but these impressions don’t generate a single install. The advertiser will not be paying anything in this situation but the publisher should still be earning $5. At scale, this is a very dangerous position for the ad-network to be in. The ad-networks today have different tools on the advertiser side to monitor fraud and traffic quality and adjust the revenue retroactively. Header bidding will require a similar set of tools to be developed on the publisher side in order to minimize risks for both sides.

Monitoring of header bidding

One area that is still unsolved for header bidding is measurement. In RTB, the SSP manages the auction on the server, collects the money and pays the publisher. In header bidding, this responsibility falls on the publisher side. The auction is managed on the client side and each bidder pays the publisher seperately based on the aggregated amount in all the bids he ended up winning. This requires a system that will billions of impressions on the client side, collect all the winning bids and aggregate them to determine how much the ad-network should be paying. Without such system, the header bidding becomes useless as it will be too exposed to abuse. At the same time, the ad-networks who are now taking the risk will want more visibility into the context in which the ads are shown and to their viewability. The requirement for better measurement will come from both sides.