Analytics

Analytics

Analytics, Industry News

IMG_4209

Last Thursday, the mobile ad-tech space was surprised to hear the news about Oracle buying mobile viewability company MOAT for 850 million dollars ($850M).  This reflects a very high multiple on revenue and a very nice return for the investors that put in an aggregated amount of $67.5M – most of it in 2016. If you have been mostly involved in mobile apps – you might have not heard about mobile viewabilty at all and are wondering how can such an unknown aspect of mobile advertising can be worth so much money to someone. To understand this – we need to dig a bit deeper.

Over 95% of the impressions are ads for installing other apps

In a recent blog post we announced a new and super exciting feature that was recently added to the Traceback platform. At SOOMLA, we are all about giving visibility to the publishers to know more about their in-app ads so knowing who is advertising in your app is a big part of that.

One of the things we quickly realized when starting to look at this data is that most of the advertising activity is driven by demand coming from other apps. These are typically CPI campaigns that only pay when the user installed the advertised app. The surprising part is that the number of ads coming from brand advertisers is very low. In the last 2 years we have seen a growing number of indications that the brands are coming to mobile.

Chart showing US mobile ad spending by industry in 2015. Retain comes first with $6.65B followed by the financial services with $3.49BIn this link you can find a report from eMarketer breaking down mobile ad spend by category. In the image to the right you can see that while retial ad spend might have a mix of brand and app install campaigns, the following categories are dominated by brand ads:

  • Financial services (Capital One, Geiko) – $3.49B
  • Automative – $3.43B
  • CPG (Procter and Gamble) – $2.33B

With at least 10 billion dollars ($10B) being spent on brand campaigns, we would have expected more of these ads to show up in mobile apps. We are not alone in our expectation of course, in 2014 Eric Seufert wrote:

“If the largest brand advertisers shift just 10% of their overall budgets to mobile, they’ll match or exceed the money spent by the app economy’s behemoths – and they’ll be competing for the same ad inventory.”

Well, according to the eMarketer report, almost 50% of the digital ad spend of these brands have shifted to mobile and yet these ads are no where to be seen. Eric was not wrong to expect a change but we all missed one important thing – the ‘other’ mobile industry.

Mobile web and mobile apps – two separate ecosystems

From a user perspective, mobile is a single experience. Opening the URL www.weather.com or opening the weather app will result in a very similar experience. The technological aspects are very different between mobile web and mobile app and each one has a separate eco-system when it comes to mobile advertising with very little overlap in between. The mobile web ecosystem was pretty much inherited from the desktop space. Each desktop advertising player in the ecosystem gradually started adding mobile web support so eventually the mobile web and desktop web ecosystems operate in a very similar way. The mobile app ecosystem evolved mainly from the need of gaming apps to acquire massive amounts of users. This resulted in an install focused industry with a great focus on attribution – a concept that mobile web companies haven’t even heard about.

What is viewability and how it evolved

In 2013, early reports started coming in showing that when brands are paying for users to watch their ads, they are often not getting what they paid for – nearly half of ads are not seen according to Comscore report from 2013. In 2014, a report from Google claimed it’s actually 56% of ads that are not viewable. This made brands worried about buying display ad inventory online and gave a big push to a category called viewability measurement. Here are some of the problems causing low viewability rate:

  • Ads can be shown In a window that is in the background and hidden by another window
  • Users sometime scroll away from the part of the page where the ad was shown
  • Time on screen is to short or video playback was stopped early
  • Traffic generated by bots rather than humans

MOAT has put it together nicely in this web image:
Non-viewable impression can be caused by 4 things: Out of focus, Out of sight, missed opportunity (area), missed opportunity (time)

The need for a solution forced the digital advertising industry to act together and set official guidelines. The Internet Advertising Bureau (IAB) together with the Media Rating Council (MRC) has published a first set of guidelines in June 30th 2014 which later evolved into version 2.0 of the guildelines in August 18th 2015.

Viewability in mobile web

As mentioned before, the mobile web ecosystem is a replica of the desktop web ecosystem in terms of advertising at least. Viewability measurement quickly became an issue in this industry as well and in June 2016, MRC published their mobile guidelines for viewability measurement. These guidelines are for mobile apps but we will touch on that later on. The main guidelines in the publication are these:

  • Client side measurement
  • Measurement of offline activity for apps
  • Filtering non-human ad views
  • Differentiate viewing from pre-rendering and pre-fetching
  • Detect when ads are out of focus
  • Pixel requirement: at least 50% of pixels were on screen
  • Time requirement: 1 continuous second of a post-rendered ad that meets the pixel requirement

The MRC is also accrediting viewability measurement companies who meet the criteria based on an annual audit.

Who can measure mobile viewability

The top 3 providers for desktop viewability measurement have all been accredited for mobile viewability measurement. Here they are:

These companies cater for both advertisers as well as publishers.
Brand advertisers are willing to pay for viewability measurement to know that their ads are getting viewed. They would typically deduct the non-viewable ads from the media campaign when calculating delivery and payout. However, most brands would also consider publishers with low viewability scores as non-safe and illegitimate media. It is common for brands to have a viewability threshold of 90% or 95% for publishers. This means that publishers with a low score will simply not get any brand ads.

On the other side, publishers also have an incentive to have their sites and apps measured. The basic reason is that if you don’t allow measurement you will not receive brand ads. On top of that, publishers who sell directly to brands or ad-networks who represent them want to know the metrics and data points about their media so they can use it in their pitching.

Mobile apps have been a slow adopter of viewability

When you check how many app publishers have a viewability measurement SDK installed you discover that most of them don’t. We can look at the 200 most downloaded apps (top free chart) on both iOS (Top chart link) and Android (Top chart link). These are typically the apps who will have advertising in them as they attract a lot of users.

  • On iOS – only 5 out of the top 200 have viewability SDK – 2.5%
  • On Android – only 21 out of the top 200 have viewability SDK – 10.5%

Our assumption is that the more you advance towards the long tail the less likely apps will have a viewability SDK since the long tail apps are still focusing on more basic aspects.

SDK fatigue slowing down adoption and enticing provider collaboration

One of the problems slowing down the adoption of viewability measurement in the app ecosystem is the need to integrate an SDK. To make things even worse, an app would need to integrate the SDKs of all 3 providers (MOAT, IAS and DV) to enjoy the full benefits. A recent open source initiative is aiming to solve at least part of the problem. It started out of IAS but was later handed over to the IAB to manage the project. The 3 providers have agreed to collaborate and support the single open source SDK that will make things a bit easier on app-publishers.
At the same time, some of the ad-networks have increased their interest level in viewability measurement as a way to be more attractive to brand campaigns. Some of them are working on bundling the viewability SDKs inside their SDKs so that their entire inventory will become available to brand campaigns. That however, brings another risk. Most apps user multiple ad-networks so if many ad-networks follow the same path an app could carry a number of viewability SDKs at the same time.

Viewability might not be enough

It’s clear that brands would not advertise in mobile apps without having viewability measurement. If you haven’t heard about the P&G $2.8B Ultimatum to the media industry – you can read about it here. However, it’s not clear if having viewability measurement is sufficient to make brands start advertising in mobile apps. The condition might be necessary but not sufficient. In other words, there could be other road blocks for brands to advertise in mobile apps. Here are some other reasons why brands could be staying away from mobile apps:

  • 50% of mobile app traffic is in games and brands have shied away from games historically
  • Brands often have issues with incentivized advertising and rewarded video falls into the category
  • The lack of audience data and tools for advertisers to target specific segments
  • Today brands are separated from mobile apps by multiple hops which takes a cut and reduces the eCPMs for the publisher so his mediation provider might not allow them to show

Oracle might be the last one laughing

So if you think of the projections for mobile viewaiblity. MOAT is already positioned as the one of the top 3 and some say the leader of mobile viewability in a market that is expected to double itself in 3 years. Considering that mobile viewability hasn’t even made it into mobile apps – the market can double in size again when Apps realize that they can get the brands competing for their inventory as well. So if MOAT can still grow 4x in 3 years, maybe the price Oracle paid is not that high after all.

Feel free to share:
Analytics, Research

Mobile Attribution 101 - the complete guide. Header banner.

For a while now the business of advertising apps have zoned in to one standard business model – cost per install (or CPI). The CPI business model has pros and cons like any other business model but it proves less fraudulent than clicks, less risky for the advertiser than CPM and more uniform compared to more advanced CPA models. With that, measuring installs and deciding which sources they came from became a highly significant part of any marketing campaign. For the advertisers, it’s not only about fair business models but also about measuring marketing effectiveness. Attribution of installs to a sources, is a key ingredient in understanding which marketing activities and campaigns are the profitable ones. For the ad-networks, being able to receive the post-backs and post-install conversion data is an important feedback loop that allows them to optimize their campaigns and maximizing yield as well as returns for the advertisers.

3rd party attribution vs. in-house

There are a few companies that tried building attribution solutions in-house. Sure, the technology is quite complex but app companies are technologically capable. Especially the gaming companies tend to prefer in-house infrastructure when it comes to analytics. The ones that did figure out how to build their own attribution solution soon realized that there is not much use for it. The most prevalent business model between app advertisers and ad-networks is CPI. Since advertisers pay the ad-networks based on installs – both sides needs to agree about the number of installs so when the advertisers build their own attribution, they soon realize that many ad-networks are not willing to trust it’s measurements. Instead, they request that a 3rd party partner will be in charge of keeping the score.

Facebook’s attempt to become a player

In 2015, in light of the growing market, mobile advertising giant, Facebook also made an attempt to become a player in this market and own a bigger piece of the end-to-end marketing solution. The market, however, rejected the move. Publisher demanded that a 3rd party will be in charge of the measurement.

The battle of the attribution providers panel at Casual connect where the only agreed upon topic was the lack of legitimacy of Facebook as an attribution provider

In the picture (credit: Venture Beat): the battle of the attribution providers panel at Casual Connect where the only agreed upon topic was the Facebook move legitimacy (or lack of).

How did this market get so big?

The 4 big attribution providers track about 18 billion dollars in ad spend worldwide. This is based on various points of information showing that Kochava measured 3 billion dollars in 2015 and Appsflyer measured 6 billion dollars in 2016. We can estimate based on the pricing of the providers that the attribution fees reached about 1% of that – 180 million dollars in 2016. The market is already big and is expected to double it’s size by 2020 reaching over $360M. Some of the driving forces behind this growth are the ad-networks themselves. Each attribution provider supports hundreds of networks and each ad-network has a lot of people in the field, meeting customers on a daily basis. Thousands of people going to every show, attending every conference and talking about CPI campaigns. Once ad-networks agreed that 3rd party attribution is a must-have, it was only a matter of time before the customer accepted that.

Attribution methods and approaches

Since the early days of attribution, there have been people saying that the models are not accurate. We live in a world where each user is being exposed to dozens of advertisements every day, often across different mediums. By the time a conversion, an app install, has occurred, the user is likely to have seen an ad for that app more than once and sometimes even as much as 10 times. Despite that, attribution models always give credit to a single source, single campaign and single ad-group. The credit is always given to the last click if such a click happened recently and if there is no click, it will be given to the last impression if it happened recently. It’s always the last one. Critics say rightfully that this is an inaccurate way to do it and there are better models. However, this is one of the areas where the delicate balance between ad-networks and advertisers mandates a simple model that is easily decide and so last-click attribution, while not perfect, is the best we have and will stay the attribution model going forward.
Last click, view thru, first click and other methods can compete but the winner is last click

The evolution of click fraud detection and IAP measurement

Being a key ingredient for measuring marketing returns or ROAS (return on ad spend). The attribution providers got pulled into providing more analytical services. The attribution dashboard became the go to screen for understanding install volumes from different sources. As the industry became more aware of the issues around traffic quality, attribution providers got pulled in to provide additional reports.
One direction in which attribution providers evolved is the area of post-install metrics such as retention tracking and conversion to payers via in-app purchases. Publishers also started using post install metrics reported by the attribution providers to set goals for user acquisition channels. It’s common these days to see an IO (Insertion Order) with traffic quality criteria. Beyond measuring the in-app purchases, attribution providers also developed the ability to distinguish real purchases from fake ones.
The other direction in which attribution developers were asked to evolve is install fraud detection and prevention. While generating fraudulent installs is much harder than clicks, publishers have a lot of motivation to do so and methods were developed to manufacture installs. All big 4 attribution providers today have developed mechanisms to detect such activity and remove it from the reports. By doing so, they are providing more value to the advertiser who only pays for real installs.

The top 4 mobile attribution providers

appsflyer logo - mobile attribution provider

Appsflyer is the most VC backed provider out of the bunch. They hold an impressive share of the top 200 companies and have the biggest penetration in far east countries. They are based in Hertzelia, Israel with offices all around the world. Their unique pricing model allows app companies to start for free which makes them highly attractive for smaller companies. Appsflyer started in the Microsoft Accelerator and is still rumored to be using Microsoft cloud infrastructure – Azure.

Name Appsflyer
Headquarters Hertzelia, Israel
Employees (by Linkedin) 208
Market Share (by Mightysignal) 12% of Top 200 Apps
Notable Customers Hulu, Cheetah Mobile, The Weather Channel
Funding raised $83M
Founded 2011
Supported ad-networks 2,148 (as of 1/1/2017)

logo image of adjust, the mobile measurement partner

The Berlin based provider has made itself a name in quite a short time. They are an official measurement provider for Facebook and pride themselves as being the leader among the top 200 apps. Adjust is leveraging their own private cloud infrastructure rather than a hosted one which also allows them to be the best at protecting user privacy.

Name Adjust
Headquarters Berlin, Germany
Employees (by Linkedin) 127
Market Share (by Mightusignal) 17% of Top 200 Apps
Notable Customers Spotify, Zynga, Rovio, Miniclip
Funding raised $29M
Founded 2012
Supported ad-networks 700+ (as of 1/1/2017)

Tune Logo - among other marketing and analytical services, tune also offers attribution solutionTune is a popular attribution choice. The company had two products going by separate brand: Has Offers and Mobile App Tracking. They are the only one of the four that is not focused completely on attribution and are also not an official Facebook measurement partner. To overcome this problem, Tune provides a solution for Facebook attribution leveraging deep linking.

Name TUNE
Headquarters Seattle, US
Employees (by Linkedin) 371
Market Share (by Mightysignal) 11% of Top 200 Apps
Notable Customers Uber, Lyft, Supercell
Funding raised $36M
Founded 2009
Supported ad-networks 1,062 (as of 1/1/2017)

kochava logo - attribution company catering to big media companies and others

Kochava is the only bootstrapped provider out of the top 4. They made themselves a name by attracting top tier media companies such as ABC, CBS and Disney as well as the mobile gaming giant – MZ. The company recently launched a new product called Kochava Collective to help their customers reach relevant audiences and also started offering a free version of their platform under the name – Free App Analytics

Name Kochava
Headquarters Sandpoint, US
Employees (by Linkedin) 82
Market Share (by Mightysignal) 11% of Top 200 Apps
Notable Customers MZ, ABC, CBS, Bigfish
Funding raised  Bootstrapped
Founded 2011
Supported ad-networks 2,800 (as of 4/15/2017)

Mid market solutions with an attribution feature

The 4 top providers mentioned above are highly focused on attribution and cater to customers who often build their in-house analytics and use the API offered by these providers to pull the data into their own BI or data warehouse system. However, there is a group of companies below the top tier that are relying on 3rd party analytical tools rather than a tool that is built in-house. These companies are preferring a full-solution approach that includes both analytics as well as attribution. This opportunity is considered the mid market of the mobile attribution space and there are companies who are starting to cater for it. Here are 2 providers in this category:

  • Tenjin – provides analytics solution with built in attribution
  • Apsalar – provides analytics, attribution and audience building under the same roof

The ability of companies to simply add attribution as a feature on top of something else is a result of the maturity of the market. The connections with the ad-networks became standardized and there are now also tools to attribute Facebook traffic with no need to get their official blessing. This is the reason why we also see companies like Singular adding attribution into their marketing and cost aggregation platform.

It’s important to note however, that these solutions are often lacking in terms of how many ad-networks accept them as an authority to counting installs. One of the key features of the attribution providers is the ability to serve as an unbiased 3rd party that is accepted by both sides when it comes to discrepancies in install counts that impact how much is paid by the advertiser to the ad-network.

Free attribution – what are the options and what’s the catch

Publishers might be surprised sometimes by the fees for the attribution service. Take Appsflyer for example – they recently tripled their price from $0.01 per install to $0.03 per install. Furthermore, if a certain app has a low volume of installs, the price is even higher. Tracking 100,000 installs for example will cost $4,000 per month. Quite a high fee for apps that are just starting out. This is part of the reason why many publishers are looking for alternative free solutions.

Option 1 – Branch Metrics deep linking

Branch offers a free deep linking solution. This means that every time an ad is shown a savvy marketer can provide a dedicated click url with additional parameters and those parameters will magically find their way to the app. Unlike other deep linking solutions, Branch takes the extra step to report these parameters back to their own dashboard and present it for the marketer to monitor the performance of each channel.

What’s the catch:

  • Ad-networks that get paid based on CPI might not agree to trust this solution
  • Advanced features like: postbacks and fraud protection are not available
  • Not scalable – every new ad requires generating additional links and passing them through to ad-networks

Option 2 – Facebook and Google attribution

For apps that only buy media on Facebook and Google, this might be a good enough solution. Both companies have an SDK that allow the app to report post install events. The impressions, clicks, installs and post-install conversions will be shown in a dashboard along side the cost. This means that marketers can calculate ROI on each campaign very easily and take action right away without leaving the dashboard.

What’s the catch: Facebook and Google are not neutral. They actually have an interest to attribute installs to themselves and given Facebook’s history with reporting errors one might be worried about putting all his faith in them. In addition, installs might be reported twice as there is no 3rd party overseeing both platforms.

Option 3 – Free App Analytics by Kochava

This is an interesting option for app publishers. Free App Analytics offers the same features as the main Kochava solution and is using the same infrastructure and reporting interfaces with one major difference – it’s completely free.

What’s the catch: Publishers most provide Kochava with a license to use the data in Kochava Collective. This means that advertisers will be able to apply more advanced targeting for these users but the ads might not necessarily appear in the app that is using the Free App Analytics solution. Might be a small price to pay for saving a few thousands of dollars a month.

In addition to these 3 options, we have also heard the names – Attriboost and DCMN as potential free attribution vendors.

How to compare between different providers

With so many options to choose from, it’s easy that some companies are finding it hard to choose. The first question to ask yourself is – do I have an in-house analytics solution. The answer to that will tell you if you are looking for a point solution for attribution or a full solution that includes both attribution, analytics and flexible visualization system that allows you to create dashboards focused on different aspects of the data.

If you want to take the scorecard approach for comparing providers that’s a good approach according to Saikala of SpaceApe. Here is her free template for comparing attribution providers with a spreadsheet. She is also giving her opinion and tips how to use the spreadsheet in this article.

Where is the attribution market headed next

One of the areas that is still unresolved is the single ROAS view. Advertisers are surprised that even after years of evolution in marketing measurement tools, there is no single provider that allows them to see a per source/campaign/ad-group view with the following components:

  • Cost of the marketing activity
  • Returns on the marketing generated via in app purchases
  • Returns on the marketing generated via in app advertising

This is the very fundamental view that allows marketers to evaluate the business merit of each marketing activity. However, today it is still very hard for marketers to get this view generated. Attribution providers realized this problem is an opportunity for them to provide a more complete solution and are now looking into two areas of expansion:
Aggregating cost data from all the different ad-networks and bringing them to the same view that summarizes the install counts and in app purchase revenues. While Singular was the first company to offer this type of aggregation, attribution providers identified the value for the marketer is far greater if he can get both elements from them rather than using an additional provider.
Assigning ad revenue to users and attributing them back to marketing activities to complete the ROAS picture. This area is gaining a lot of momentum as apps are expected to increasingly rely on ad-revenue according to industry forecasts. The first company to provide a solution in this field is SOOMLA but attribution providers are now working to add these capabilities by partnering or building in-house.

Feel free to share:
Analytics, Marketing

blog header image with postbacks for ad whales written as the title and signs that say where, what, why, who, when and how

Before we jump into the topic of postbacks for ad whales, lets first understand what are postbacks and why they are so important for any marketer of a mobile app company. Let’s say you have a dating app called TrueMatch and after you have had some organic growth you have recently partnered with a few marketing partners – mostly ad-networks who specialize in bringing installs. Let’s call one of them Tap4Buck. Tap4Buck places ads to promote your app TrueMatch on different websites and apps. As a result users click on them and get to your app-landing page. Some of them also decide to install your app and a smaller percentage even continues and converts to payers. Since Tap4Buck wants to give you the best results possible, they want to know which clicks ended up converting to installs and which ones converted to purchasing users. The problem is that the app store landing page breaks the flow of information so Tap4Buck can’t continue to track the user once they have installed the app. Postbacks solve this issue. If you are using an attribution provider (you should – it’s a must have these days), you can easily configure it to send postbacks to Tap4Buck and help them optimize your campaign for you.

What are ad whales and what are postbacks for ad whales

Now, let’s imagine that TrueMatch makes 50% of it’s ad-revenue from advertising. This means that sending postbacks for users who made purchases only tells Tap4Buck half the story. What about users who generate a lot of revenue from ad based monetization? Ad Whales are users who made at least $0.7 in ad revenue. This is the minimal amount of revenue a payer can make ($1 purchase minus 30% cut by Apple/Google). $0.7 threshold means that a conversion to ad-whale yields the same amount of money as a conversion to payer would yield. Postbacks for ad whales means that your attribution provider would send Tap4Buck an event every time a user that came through Tap4Buck has generated at least $0.7 in ad revenue and converted into an ad whale. This typically happens with 2%-5% of the users in games that are tuned towards ad based monetization but obviously changes from one game to another.

Who should care about postbacks for ad whales?

Companies who have any type of paid marketing activity would benefit from sending postbacks in general. The ones that also have an ad revenue component that amounts to at least 15% of their total revenue should be sending postbacks for ad whales. Ad whale postbacks also benefit the partners on both sides. For the marketing partner that sent the traffic to your app, better postbacks means more effective campaigns and happier customers. For the monetization partners, better postbacks means that the app will get more ad whales as a result of the optimization and therefore their volume of revenue would increase.

When – 2017 is the year of change

If you have been following the industry trends you already know that ad revenues are becoming the dominant way to monetize apps. It’s already as big as In-App Purchase and is projected to grow faster in the next 4 years. In Mobile games alone, App Annie projects in-app advertising will amount to revenue streams of over fifty billion dollars ($50B) for the companies who will be placing these ads in their apps. The total mobile ad spend worldwide is projected to reach $195B by eMarketer. As ad based monetization is becoming so important, companies are looking for tools to optimize them and postbacks are a big part how the mobile marketing space has been operating.

Where – not all geographic areas are created equally

Most of the media buying today is concentrated in a few countries where people are willing to spend money on in-app items. These countries are often referred to as Tier 1 countries and are also where most of the postbacks are being fired today. At the same time, postbacks for ad whales bring a new opportunity to table. There are other countries with large population where people can’t afford to buy in-app items. These countries offer low rates for user acquisition due to lack of demand. Setting up postbacks for ad whales allow app publishers to find opportunities to acquire users in these countries with positive ROI. This means that as postbacks for ad whales became more popular through out 2017 we will see a shift in the postback geographical activity areas.

Why track conversion to ad whales and report it as postbacks?

There are 3 main reasons to track and post ad whale conversion:
Business goals alignment – many apps that have a big ratio of ad revenue today would make up a game progress goal such as “100 sessions completed” or “10 levels”. These goals would be defined as events and companies would track conversion to these goals and report postbacks to the ad-networks. However, these goals are not aligned with the business of the company. Conversion to payers and to ad whales is a far better goal and will bring better results in the long term.

ROAS not enough
– Measuring and optimizing the return on ad spend is the best theoretical approach. However, in a real world situation it relays on predictive models that are often hard to implement. Media buyers often require a more day to day metric to optimize against. This is why most UA campaigns track the conversion to payers as one of the leading KPIs. Similarly, in apps that monetize mainly with ads, the easiest way for media buyers to optimize is against a goal of conversion to ad-whales.

Postbacks allows manual as well as automatic optimizations – reporting the conversion to ad whales as a postback to the traffic source allows them to have an optimization goal that is aligned with your business. In turn, it impacts what users you will be getting from this traffic source. In some channels such as search and social media there is a lot of algorithmic optimization taking place. These algorithms need a goal to optimize against so having them optimize for ad whales would be the best approach for an ad supported app. Similarly, in other channels there is a manual optimization process of eliminating bad sub-sources such as sites or segments – these manual optimizations also requires a goal and reporting ad-whale conversion as postbacks provides such a goal.

How to set up ad whale conversion as postbacks

There are 3 components for setting up ad whale postbacks in your app:

#1 – Tracing back ad revenue per user – in order to detect the ad whales and report them you will need a way to measure the ad revenue for each user separately. Your monetization partners typically report ad revenue per country and average CPM but not the ad revenue for specific users. The most accurate way to measure ad revenue today is SOOMLA TRACEBACK. It is the only platform that can identify the ad whales for you.

#2 – Connecting the data pipelines – your attribution platform is the one in charge of sending postbacks to your marketing partners. Once you have SOOMLA integrated in your app you can configure it to send the right postbacks to your attribution platform with just a few clicks.

#3 – Setting up postbacks in your attribution platform – this step is slightly different depending on the attribution partner. However, they all have a partner configuration screen where you can set up the ad-whale conversion from phase #2 as the trigger for the postback.

 

Feel free to share:
Analytics, Marketing

Reality can prove very different than the statistics that represent it

There is a simple idea at the core of most mobile marketing campaigns these days – if you spend $x on some marketing activity and received $y in return you want y to be grater than x. This is often referred to as ROAS or campaign ROI. We have trained mobile marketers to break down their activities to small units: ad groups, ad sets, ad creatives, audiences, … and find the ones that show ROAS. Doubling down on the positive ROAS units while shutting down the negetive ROAS units is the leading campaign optimization strategy today.

Here is the problem – it only works under certain conditions.

There is a famous saying by Mark Twain – “There are lies, damned lies and Statistics”. It comes to warn people about using statistics in a wrong way. One such way is using statistics when small numbers are involved. Another way in which statistics are deceiving is called Multiplicity or Multiple comparisons. Let’s see how those come into play when calculating returns.

Beware of the small numbers

Most companies base their ROAS calculations only on revenues from In-App Purchases. This is a result of 2 things:

  • Up until recently, ad based monetization and ad spend were mutually exclusive
  • Until SOOMLA TRACEBACK there was no way to attribute ad monetization

The problem with In-App Purchases revenue is that it’s highly concentrated. Studies have shown that purchases are less than 2% of the users and among those 2%, the top 10% generate half of the revenue. Let’s say that you spent $5,000 to acquire 1,000 users and you are trying to figure out the return. Most likely you have 20 purchases but there are 2 whale users who generated $1,500 each (this is aligned with the studies – yes). Now, suppose you had 2 ad-groups in that campaign and you are trying to figure out which one was better. Here are the options:

  • Group A had both whales
  • Group A had one whale and B had one whale
  • Group B had both whales

Since we are talking about 2 users here – the scenario that actually happened would be completely random. Even if one ad-group is better than the other it is still very likely for that group to outperform the other group when we are talking about only 2 users who can flip the outcome completely. The danger here is that our UA teams would double down on the ad-group that yielded the 2 whales without understanding that it’s not better than the other. If we look at sample sizes here n=1000 is normally considered a good sample size. Has the monetization been less concentrated a sample size of 1,000 should have been enough to make decisions. However, for the purpose of acquiring whales the actual sample size is n=2 in this case. We should try to get at least n=500 before we start making decisions on media buying. The problem of course is that attracting 500 whales could be a very expensive test – more than $100,000 based on the numbers in the example above.
On the other hand, companies who monetize with ads enjoy the fact that more users participate in generating revenue and can make decisions based on smaller sample sizes and smaller test budgets.

Multiplicity – the bias of multiple shots

Another bias we normally see in mobile marketing is Multiplicity. The easiest way to explain this is with the game of basketball. Let’s imagine you are through from the 3-pt line and you have 50% chance to score. What happense if you try twice, the chances of scoring at least once becomes 75%. With 3 shots, it’s 87.5% and so forth. The more times you try the better your chances to score.This is what happens when you try to hard to find positive ROAS in a campaigns that has a lot of parameter. You compare ad-groups – that’s 1 shot, you compare ad creatives – that’s a 2nd shot, you compare audiences – that’s a 3rd shot and so forth. The more you try to find a segments with positive ROAS by slicing and dicing the more likely you are to find a false positive one.

Feel free to share:
Analytics, App Monetization

Ad revenue concentration hero image with chart and text

We are happy to report some interesting data points we recently looked at. The goal was to understand how concentrated ad revenue really is. Everybody knows already that the 80/20 law applies in IAP – at least 80% of the revenue is driven by the top 20% of purchasers. There is plenty of research showing how concentrated IAP revenue is. Ad revenue, however, is still a mystery for most publishers and very few companies actually have the data on how concentrated the revenue is. If you take the naive approach and believe all the users contribute revenue based on the average eCPM you might think that the ad revenue concentration chart will be flat. The reality however, is very different.

Comparing Concentration in Ad Revenue vs IAP Revenue

In the image below you can see a comparison of the revenue concentration between ad based monetization and IAP based monetization. These charts are based on data from 28 days of activity in a Match-3 game where most of the monetization comes from interstitials. The revenue model behind the ad monetization is CPC in this case.

On the left side, the IAP revenue is highly concentrated and 80% of the revenue is generated by the top 20% of the users. The top user generated more than $300 in revenues for the app.

On the right side, we see that the ad revenue is also highly concentrated. The top 20% are contributing more than 50% of the revenue here and the top user generates $2.5 while there are users who only contribute a few cents.

Comparison between concentration of ad revenue and IAP revenue

Ad Revenue Concentration with Reward Ads

One of the hot trends of 2015 and 2016 was the adoption of rewarded video ads by many game publishers. We wanted to look at the ad revenue concentration in rewarded video as well. The chart below does exactly that.

The data here is from a single day so obviously more concentrated than information aggregated over an entire month. The game here is an mid-core action game and the monetization is done with both rewarded videos and an offer wall.

The ad revenue concentration is much higher in this data set. The top 20% of the users are contributing 90% of the revenue and the top user is contributing more than $15.

Ad revenue concentration in rewarded video ads and offer walls

Who are the users contributing high amounts of ad revenue

Once you realize that the ad revenue is concentrated almost as much as iAP revenue, your next question is likely to be: “how are these users”. On a high level, these users are typically the users who download many apps as indicated by a Comscore Report highlighted in this article. But you can go a lot further than that. Using SOOMLA Traceback you can profile these “Ad Whales and target them in marketing activities.

 

Feel free to share:
Analytics, App Monetization

Reward Abusers written in blue text next to a trophy and Heavy App Downloaders written in green text next to an app icon with two axes

As the market adopts TRACEBACK technology we are learning new things about how users interact with ads. This allows us to classify users into types. Let’s think about these two types of users who are highly relevant to rewarded video based monetization.

Reward Abusers – these are users who watch the videos to get the rewards but are not contributing any revenue in Neither IAP nor in Ad revenue.

Heavy App Downloaders – these are users who download and try multiple apps each month. Typically, these are the users who end up generating the most amount of video ad revenue for your app.

How these segments impact your business

Let’s say you are buying traffic from a new source. You probably ask yourself, how many installs I received but you should also ask the million dollar question – “what type of users am I getting?”

Why?

Consider 2 possible sources:
Incent Campaign – this campaign gives users an incentive in another app in return for downloading your app. By nature these users are after the rewards so this source might be heavy in Reward Abusers
FB Campaign – now consider a campaign targeting lookalikes of users who are existing Heavy App Downloaders. This campaign is likely to bring more Heavy App Downloaders. You can learn more about this specific technique – here

How can you segment your users

If you are already convinced that knowing the Reward Abusers from the Heavy App Downloaders can impact your business your next question should be how to spot them. Let’s think about what features are similar and which ones are different between thm.
App Engagement – both user types have high app engagement
Video Ad Engagement – Reward Abusers will watch as much videos as Heavy App Downloaders
Post Impression Performance – this is the feature that sets them apart – Reward Abusers will only watch the videos while Heavy App Downloaders will also click and install the apps presented to them

Reward Abusers Heavy App Downloaders
App Engagement High High
Video Ad Engagement High High
Post Impression Performance Low High

So understanding what the user does after he watches the video ad is the key here. Today, there are two solutions in the market:
Developing In-house – this requires your engineering team to figure out specific ways to track post impression events with each ad-network and to keep updating the code every time there is an update to the ad-network SDKs
SOOMLA TRACEBACK – our platform does all the work for you, it requires a simple integration but once implemented you will be able to segment your users reliably, track ROAS and do many other mind blowing optimizations to your ad-revenue. CLICK TO LEARN MORE

Feel free to share:
Analytics, App Monetization

Podcast - The role of analytics and data in ad based monetizationThis is a recent interview I gave at Cranberry Radio. I’m talking about the following topics:

  • The role of unbiased measurement companies on the advertiser side and the publisher side
  • Insights that SOOMLA have seen by measuring ad-revenue to an unprecedented granularity level
  • Insights from our latest study – Rewarded Video Ads Retention Impact in Match 3 Games

You can listen to the podcast here.

Feel free to share:
Analytics, Marketing

4 top mobile a/b testing tools header image

This is a guest post by Natalia Yakavenka from SplitMetrics

Ask any mobile marketer what is the best way to optimize conversion rates for your app page and you’ll most likely get A/B testing as a response. While A/B testing is still most often associated with the web, the concept of a/b testing for mobile app pages is not new. The very first solutions growth hackers used were custom coded landings, but such approach requires time and effort. However, app page conversion optimization only became popular when self-service platforms like SplitMetrics and Storemaven emerged. These platforms brought a completely new level of A/B testing for mobile pages as they provided insights on top of showing the winning variation. Later on, the introduction of Google Play Experiments in 2015, brought A/B testing of app landing pages into the “must have” category for app marketers. Since that time, plenty of new solutions have emerged but we recommend sticking to the 4 most popular tools presented here.

Google play store allows experiments - a limited way to do mobile a/b testsGoogle Play Experiments

When it comes to selecting the best A/B testing tool, the most common question is why go elsewhere if you have the free Google Play Experiments. Indeed, it allows mobile publishers to run free experiments on their app pages, but it comes with significant limitations. The most serious ones are that you can’t test unpublished apps and you’ll never find exactly what worked due to the lack of on-page analytics. Still, Google Play is the perfect solution for those who are not familiar with paid traffic and user acquisition as it doesn’t require driving traffic to the experiment from ad resources. The other three tools require sending traffic to their experiments and are usually for more advanced marketers.        

Distinctive features: absolutely free + requires no additional traffic

Split metrics logo - this mobile a/b testing tool offers many advanced ASO featuresSplitMetrics    

Founded in 2014, SplitMetrics was among the first ones to provide every marketer with an easy-to-use, unlimited, and flexible A/B testing tool. In addition to the regular icon/screenshot testing, it offers pre-launch experiments for unpublished apps and Search, Category and App Store Search Ads testing. Unlike the Google Play service, it offers a multi-armed bandit approach which helps reach significant results fast. But it’s not as ideal as it seems to be — you have to pay for it. Though the price is very reasonable and you have a 30-day trial, you will need to pay a monthly fee for your subscription.

Distinctive feature:  pre-launch experiments and App Store Search Ads testing

StoreMaven is one of the pioneers of the mobile a/b testing and ASO spaceStoremaven

StoreMaven provides easy-to-use A/B testing for the entire app store landing page experience. One of their advantages is offering benchmarks based and best practices based on their broad client base in each of the app store categories. On top of that, StoreMaven clients benefit from their money saving algorithm, StoreIQ. This algorithm helps conclude tests with fewer samples and lower costs by leveraging historical data to quickly determine the winning creatives. StoreMaven provides a fully dedicated Account Manager to make sure clients make the most of their testing budgets. This tool is also a paid tool that is offered as a monthly subscription.

Distinctive feature: Professional Services

One of the features offered by the tune platform is A/B testing to improve your ASOTUNE’s A/B Testing

4) Tune offers many services for app marketers. They are mostly known for their attribution service — measuring paid app installs. However, they also offer A/B testing and optimization tool for the app landing page. Launched in spring 2016, it already provides solid functionality, offering the basic functions of testing different types of assets and showing all measurements and stats collection. While Tune offering is more complete compared to the other testing tools, it’s biggest limitation is that it doesn’t work with other attribution providers. The tool is also limited with regards non-US regions and only supports a small list of regions. In terms of pricing, Tune’s A/B testing tool is not available as a stand alone and so customers have to buy it as part of a suite of services.

Distinctive feature: works very well with other Tune capabilities

A/B testing can be easy with the right tools and is recommended for any app marketer as part of a data-centric growth strategy. Feel free to also try our quiz — test yourself to see how data-driven your game is.

Feel free to share:
Analytics, App Monetization

Header image showing how complex it is to a/b test your ad based app monetization

A/B testing has been an integral part of marketer toolbox for a good reason – it takes a great deal of the guess work away from marketing. In online and mobile companies it also became a popular tool for product managers. Every time a new version is released, why not a/b test against the existing version and make sure nothing got broken. In mobile app monetization, however, this tool is not available.

Why ad based app monetization is so hard to A/B test

The core requirement for A/B testing is to be able split your users into two groups, give each group a different experience and measure the performance of each one so you can compare it later. There are a number of tools who can facilitate the split for you including Google Staged Rollout. If you are measuring IAP monetization it’s easy enough to associate purchases to the users who made them and then sum the revenue in Group A and Group B. In ad monetization however, it’s impossible to associate ad revenue to individual users. The ad partners mostly don’t report the revenue in this level of granularity.

Method 1 – interval testing

One alternative that companies have been using is interval testing. In this method, the app publisher will have one version of the app already published and will roll out a version with the new feature to all the devices. To make sure all the users received the new version publishers will normally use force update method that gives the user no choice. The impact of the new feature will be measured by comparing the results over two different time intervals. For example, Week1 might have contained version 1 and week 2 might contain version 2 so a publisher can compare version 1 vs. version 2 by comparing the results in different date ranges.

Pros

  • Very simple to implement – no engineering effort

Cons

  • Highly inaacurate and subject to seasonality
  • Force update method has a negative impact on retention

Method 2 – using placements or different app keys

This is a pretty clever workaround for the problem. Most ad providers has a concept of placements. In some cases, they are called zones or areas but all 3 have the same use – they are planned so you can identify different areas in your app where ads are shown for reporting and optimization purposes. The way to use this for A/B testing is to create a zone A and Zone B and then report Zone B for users that received the new feature while reporting Zone A for the control group. If you are already using the zones feature for it’s original purpose, you might already have zone 1, 2, 3, 4 and 5 so you would create 1a, 1b, 2a, 2b, ….

Of course, if you are using multiple ad-networks you would need to repeat this set up for every ad-network and after the test period aggregate the results back to conclude your A/B test.

A variation of this method is to create a new app in your ad-network configuration screen. This means you will have 2 app keys and can implement one app key in group A and the other app key in group B.

Pros

  • More accurate compared to other methods

Cons

  • The effort for implementing a single test is very high and requires engineering effort
  • Will be hard to foster a culture of testing and being data driven

Method 3 – counting Impressions

This method requires some engineering effort to set up – every time an impression is served the publisher reports an event to his own servers. In addition, the publishers sets up a daily routine that queries the reporting API of each ad-network and extracts the eCPM per country. This information is than merged in the publisher database so that for every user the impression count for every ad-network is multiplied by the daily average eCPM of that ad-network in that country. The result is the (highly inaccurate estimation of the) ad revenue of that user in that day. Once you have this system in place, you can implement A/B tests, split the users to testing groups and than get the average revenue per user in each group.

Pros

  • After the initial set up there is no engineering effort per test

Cons

  • Settting this system up is complex and requires a big engineering effort
  • Highly inaacurate – it uses average eCPM while eCPM variance is very high
  • Can lead to wrong decisions

Method 4 – leveraging true eCPM

This method leverages multiple data sources to triangulate the eCPM of every single impression. It requires significant engineering effort or a 3rd party tool like SOOMLA TRACEBACK. Once the integration of the data to the company database is completed, publishers can implement a/b tests and can get the results directly to their own BI or view them through the dashboard of the 3rd party tool. Implementing A/B tests becomes easy and a testing and optimization culture can be established.

Pros

  • The most accurate method
  • Low effort for testing allows for establishing a testing culture
  • Improvement in revenue can be in millions of dollars

Cons

  • The 3rd party tool can be expensive but there is usually very quick ROI

 

Feel free to share:

Join 7331 other smart people who get email updates for free!

We don't spam!

Unsubscribe any time

Categories

SOOMLA - An In-app Purchase Store and Virtual Goods Economy Solution for Mobile Game Developers of Free to Play Games