GameChangerSF Mobile Marketing Blog

Develop Your Own Market Research with Facebook Ads

When developing a mobile app, there are countless decisions that need to be made early on. A big part of the decision making process is figuring out the theme and branding. Building the right identity for an app is one of the most important components and companies spend big dollars making sure the brand is on target and easy for consumers to grasp. For indie game developers or smaller mobile game companies, the luxury of big budget market research is non existent. Knowing how to leverage existing tools to get consumer insight is possible, you just need to be creative.

The core strategy of online advertising is A/B testing. This allows advertisers to frequently examine what type of messaging works best for their target audiences. So why not apply this concept to testing game elements such as working titles, new characters, or new game features?

We recently had the opportunity to launch a market research ad campaign with one of our clients, Behaviour Interactive. The challenge was to find a name best suited their mobile puzzle game Home. Our strategy was to launch a series of mobile install ads targeting women 30-55 years old, featuring creative displaying the characters and a game name to test. 

To break it down, each campaign at the highest level was clearly titled by the objective. At the next level each ad set is titled by the test subject (i.e. game title), followed by the ad which features the game name being tested. The ad that receives the highest click-through rate and the most conversions is theoretically the preferred name for the app. Here’s an example of how this is structured on Facebook. 

After we ran several rounds of testing for Home, we had enough data to determine the best performing name, “Home Pop!”, showing us that the target audience was more receptive to a short, simple and fun name. We also saw a statistical tie between all but one of the ads based on installs and CTRs, showing one to be inferior. 

Some of the benefits of doing this type of testing include:

 • Can provide data to help make key decisions in a games evolution
 • Targeting a sample group made up of a mobile game’s primary audience
 • Campaigns can be quickly launched with short runs
 • Cost is minimal as ad platforms such as facebook allow advertisers to spend a minimum of $1 to run ads

Bonus Update: Recently Facebook launched a new feature allowing advertisers to divide target audiences into smaller sample groups for optimal testing of creative, bids, etc. without worrying about one ad affecting another targeting the same audience.

To set up User Segments for an experiment you’ll need to:

1. Create a campaign 
2. Create separate ad sets for each creative or ad variation within that campaign
3. Select a User Segment Custer from the Categories drop down
4. Run all ad sets at the same time period and compare results

Planning on running a test of your own? Got questions about Facebook’s User Segments? Drop us a note at or on Twitter @Gamechangersf
Read more ...

How to Use Facebook Targeting for Mobile Marketing

Facebook is one of the most powerful platforms for mobile games as it offers a plethora of ways to target mobile users. Compared to other social networks, Facebook is miles ahead with one of the most sophisticated advertising platforms available. Knowing how to use their ad features is essential for getting the most out of your mobile marketing campaigns. Here are some key targeting options that can make your mobile marketing plan kick ass. 

Behaviors - This feature gives advertisers the power to target users by actions they have taken on Facebook. Some examples of behaviors that can be targeted include (but not limited to):
  • Frequent travelers 
  • Small business owners 
  • Photo uploaders 
  • Event creators 
  • mobile users per device 
  • Online spenders 
  • New mobile and tablet users 
  • Car owners per brand 
  • People who’ve played a game recently (this is made up of games that have installed Facebook SDK) 

Broad categories - This feature allows advertisers to target based on different categories that users fall into based on their actions on Facebook. Some of these categories include engaged adventure, strategy, puzzle, casino, and other types of gamers.

Custom audiences - This is one of Facebook’s most impressive targeting options. With custom audiences, advertisers can target users who are already familiar with the brand. This includes people who’ve visited your site or people who’ve used your mobile app. This feature is perfect for retargeting and retention campaigns.

Lookalike audiences - Taking custom audiences a step further, lookalike audiences allows advertisers to reach new users who are likely interested in your app or business because they're similar to customers in your custom audience. Essentially you’re creating a clone of your custom audience with new users who are similar. This type of targeting delivers one of the highest click through rates overall for mobile ads.

Interests - This targeting feature allows advertisers to target specific products including mobile games. This is a great way to target competitor games that people have played or mentioned in their Facebook accounts. Interests can help build the right audience for new mobile game IPs.

Do you have any questions on mobile marketing? Drop us a note at or on Twittter @Gamechangersf

- Nelly
Read more ...

Mobile Ad Targeting Gone Crazy Bad

Awkward conversations about mobile ads 

My 7 year old daughter is not likely to respond to the survey on gun control ads that were placed in her game. The advertiser wasted their money.

How to avoid throwing away your mobile dollars

It's actually pretty simple. Before placing ads with any network, in addition to the typical questions about pricing, ask about how they target and optimize their ads. Some networks, like Facebook, let you target based on demographics, interest, geography, while networks like Chartboost will let you know exactly what apps your ad will run in and will let you exclude make-up games from your gun control campaign.

But the current problem in mobile advertising is that networks mask the name of the apps in which you are running ads. So the above app was probably not targeted by name but appeared as something like "app id 33244" and the advertiser had no idea this was going to happen. The unfortunate thing about this incident is that likely some combination of these things happened:
  1. Ad network  didn't provide transparency into placements (or advertiser didn't bother to look)
  2. Ad network didn't screen the ads and match it with appropriate apps
  3. Ad network lacked tools for app owner to block inappropriate and non-relevant ads (or app owner didn't bother to look)
Before placing ads you want to spot ad networks and situations where this can happen. Whenever possible, keep your buys to networks that have targeting. It's worth the higher cost to actually target ... your target audience.

Our industry is still young, and this is expected to happen, but with diligence and knowledge we can make things better.
Read more ...

Crunching Chartboost's CPI Data

As a mobile game marketer I am often asked about the trend in the cost of user acquisition for games and how it's possible to stay profitable. There are many articles and even data out there about the rising cost of installs (CPI). However, savvy mobile performance marketers know that the picture is more nuanced and with rigorous optimization, using relevant data, profitable user acquisition is possible.   In order to do it, you have have a strong knowledge of your app's key performance indicators (KPIs) by territory and by device.

Knowing the lifetime of value (LTV) of your players and CPI trends by territory and platform can open new possibilities for marketers. At GameChangerSF we closely monitor our clients's KPIs and are always expanding geographic reach with an eye towards increasing ROI.

Chartboost's Treasure Trove of Data

One of the public sources of data I frequently look at is Chartboost's excellent monthly CPI by device and country report.  I asked to get their historical data and they agreed. Thanks Chartboost! 

Before I present the data, I want to explain why I will take Chartboost's data over many other sources:

Chartboost shares our belief in transparency. They are one of the few networks that show you exactly where you are advertising and lets you white and black apps in which your ads appear. This leads better targeting and more efficient optimization.

Scale and Scope:
Charboost is one of the top networks and has scale in many territories. They also are one of the few networks  that covers iOS, Android and Amazon's Kindle devices. However, I do not know the number of data points in each territory/device dataset but I am sure there are places where its thin (i.e. Zimbabwe, Amazon in September 2013) so please keep that in mind. The dataset is very solid in major territories.

Focus in Gaming:
Gaming apps have different CPIs than other apps, Chartboost's focus in gaming makes this data more relevant to gaming apps.

Quality Ad Inventory: 
Finally and, most importantly, Chartboost's inventory is non-incentivised and only includes interstitial and video ads. In our experience these are the creatives that convert best. Most reports I have seen mix incent and non-incent inventory as well as lower quality creative. 

Without further ado,  I hope you enjoy these visualizations. And as always, please post any questions or comments or contact me at

Read more ...

Successfully Launching an Android App in Google Play

On April 12th +Anya Shapina and +Phil Shpilberg shared strategies for successfully launching apps on +Google Play  at +BeMyApp's AndroidFest.

The following is the video and presentation from the event. Enjoy and send us any questions at or post in the comments.

Read more ...

Making Your Toy: A Toy-Based Approach to Game Design

(Photo Credit: DaMongMan)

You think you have an idea for a really cool game design. However, you are not sure how to approach it. Do you work on the story of the game? Or perhaps you want to focus on the characters? Maybe it would help if you think of yourself as a toy designer.

What is a toy?

Before you can make the toys of your game, you must figure out the definition of a toy. It is simple to think of a toy as an object which you use for play. For baseball, the toys include the ball and bat. In Monopoly, the toys consists of the various pieces that you use to interact with the game. Within the context of video games, the toys can include the actual characters themselves. For example, Pac-Man or Chris Redfield from the Resident Evil games would all be toys since you use them for play.

What does my toy do?

The next thing you can do is think of the properties of your toy. In other words, how do people play with this toy? 

When you are designing a toy for your game, try to think of some fitting verbs that you can use to describe the actions of your toy. It is a good idea to list many action verbs and to narrow down the ones that you think might be the most interesting when playing your game. If you want something interesting, try to think of new verbs that your toy can perform. For example, the verbs "jump" and "shoot" are overdone in video games.

Here are a few toys, the games that they are associated with, and their action verbs from which you can derive inspiration from when designing your own toy:

Toy: Mario
Game: Super Mario Bros. (1985)
Run, Jump, Shoot, Climb (vines), Swim, Crouch, Stomp

Toy: Tetromino
Game: Tetris (1984)
Spin, Fall, Clear (Line)

The Flock (from right to left): The Blues, Red, Bomb, Chuck, and Matilda

Toys: The Flock
Game: Angry Birds (2009)
Fling, Crash, Split (The Blues), Scream (Red), Explode (Bomb), Dash (Chuck), and Drop Bomb (Matilda)

How would each of these games be different if the verbs were changed? For example, what if Mario tried to make peace with his enemies instead of stomp them? Try to think of how your action verbs will influence the aesthetics (the player's emotional response) and dynamics (the run-time behavior that interacts with the player's inputs and all outputs) of your game. 

The entire aesthetic and dynamic of Super Mario Bros. would change if Mario sought peace instead of destroying his enemies. The game might not be as action-packed or frenetic if this approach was applied. However, what if Mario had to make peace while also avoiding death? Worse, what if he had to resist the temptation to kill his enemies at all cost? 

Metal Gear Solid 3 comes somewhat close to this imaginary scenario. The game allows you to use Naked Snake to either kill or avoid your enemies. However, you are punished much later in the game with a difficult boss fight if you decide to kill your enemies. The procedural rhetoric of the game suggests that murder carries a price.

The MDA Framework is a tool that helps you create the appropriate emotional responses to the games that you design and develop. It will give you a good idea on how to balance your action verbs with the appropriate emotional response that you want to create in your players. It is worth considering new ways to use your toys to engage people's emotions. However, if you desire to make money in this industry, you have to validate and test what people find fun. Games such as Zynga's Farmville reveal that a large segment of people are content with managing virtual farms using verbs such as "plant", "water", and "harvest". What are some interesting ways that you can use your toys to hook people's emotions?

What now?

Creating toys and their associated action verbs is just one step in the process of making a game. However, it is important to decide early on how your toys will perform before you make the rest of your game. You have to figure out the toys in your game, list the appropriate action verbs, validate and test your toys within a game prototype, and figure out ways to monetize your game. This model of game design can be applied to a number of different sectors in the gaming industry, including console gaming, PC gaming, and even mobile gaming.

Derek Rumpler is a mobile acquisition intern with Game Changer and an indie game designer. He was producer for the award-winning Android game Edgar Allan Poe's The Tell-Tale Heart by Macabros Games. He will receive his degree in Interactive Media Studies from Miami University in May 2014. You can follow him on Twitter: @DerekRumpler.
Read more ...

The Mobile Blitz Campaign Sugar High Addiction

Blitz Campaigning (otherwise known as "burst campaigning") is driving a lot of incentivized installs (installs by people who are incentivized with a reward to download and open an app) in a short period in order to drive the app up the app store charts, and get a lot of organic installs... or so the theory goes.

A blitz campaign in the US typically lasts two to four days over a weekend and costs between fifty to over one hundred thousand dollars. Once the budget is exhausted and the chart position is reached, masses of people are expected to discover the app and install it. Expected… but let me share a secret: it rarely happens.

In a more typical scenario, organic installs fail to show up sending the app in a tailspin down the chart. Users acquired cheaply as incentivized installs don’t come back leaving the buyer with nothing to show for the spent marketing budget.

In this respect, blitz campaigns can be compared to Sugar High that one gets after consuming a large amount of sugar. One briefly feels happy and energized, but very quickly mood and energy level drops, leaving one feeling drained and down.

Incentivized Installs are at the core of the blitz strategy, and they are much like Sugar:

Non Essential
Sugar is not essential for your nutrition (glucose is). Sugar calories are “empty” calories.

Incentivized installs are not essential for your growth strategy. Incentivized installs are “empty” installs that deliver zero value. Incentivized users do not engage with the app, do not come back and certainly do not monetize. Moreover, most of them un-installs your app upon collecting their reward

False Claims
Food manufacturers will tell you that agave, honey and other sweet products are better sugars than pure sugar. Maybe a tiny bit better, but they are still essentially sugar and you will not get your nutrition there.

Incentivized installs
Companies that sell incentivized installs will tell you how theirs are better than the other guy’s. Not true.  Maybe some incentivized users have return rate have return rate of a couple percent versus zero percent from another source, but it makes no difference.

Sugar is addicting. Although the addictive nature of sugar is milder compared to drugs, eating sugar every day and multiple times each day can definitely turn into a real addiction.

Incentivized installs
Driving volume with mostly incentivized installs is an oddly habit-forming activity. Low CPIs and high download numbers are so seductive that many executive teams close their eyes on the fact that retention plummets and they contribute nothing to the actual growth (as growth requires good “stickiness”).

The Budget Dilemma
Many families in poor communities cannot afford fresh produce and instead rely on processed junk food and sugary sodas.

Incentivized installs
Many developers fond of blitz campaigns think that they simply cannot afford paying over $2 for high quality installs when their LTV is around or below $1. Growing your business by a series of burst campaigns when you can’t afford quality acquisition is like putting your kids on a diet of sugary drinks because you cannot afford fresh produce and vegetables. You have to stop and ask yourself: am I doing more harm than good? If your LTV is low, the last thing you should be doing is taking chances with your limited marketing budget.

Ditch It
If you want to lose extra weight, be healthier and have good teeth, eliminate sugar from your diet.

Incentivized installs
If you are looking to improve lifetime value of your users, if you want to generate positive return on your marketing, you need a clear picture of what your users are worth. Go on marketing cleanse: start with eliminating incentivized traffic from your acquisition mix to get a true picture of your KPIs. Fix your retention by eliminating the user drop off points – you’ll need good analytics for this. Find lower cost sources of quality installs like direct deals. Get creative and feel free to talk to us about it.

Is a Blitz Campaign Ever Worth It?

So, when is a blitz campaign justified? Surely some of them must succeed or otherwise why would anyone do it?

Many factors need to align in order to create favorable conditions for a blitz.  First, it must be an iOS app. Next, the app has to have mass appeal; a niche app will generate low to no organic installs no matter how high it is up the charts. Second, the retention has to be good in order for the new installs to translate to a lasting increase in DAU. The monetization has to be good enough to generate a positive return on the campaign. If your customer lifetime value is below $2, it would be very difficult to achieve even with a good organic rate during the burst.  Finally, you much have an attribution solution (like MobileAppTracking, Ad-X or Kochava) in order to be able to work with multiple ad partners without integrating many SDKs, track paid and organic installs and avoid paying for duplicates. And even then there is no guarantee that your blitz campaign will succeed.

Oh, and did I mention that blitz campaign is a totally losing strategy for Android apps? Google algorithm punishes for low engagement and un-installs which is what you get with incentivized traffic. 
Try it and you will get the Googleplex-suplex (like a regular suplex it will hurt, but much nerdier).

To simplify the matter, here is a decision diagram on bursting:

Read more ...