Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
September 26, 2017
arrowPress Releases






If you enjoy reading this site, you might also want to check out these UBM Tech sites:


 

Why you should Soft Launch your Mobile Game

by Niel Dagondon on 06/24/16 10:21:00 am

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.

 

Soft launching is easy but doing it right takes a serious amount of time and effort. I have seen far too many developers just soft launching because everyone else is doing it - but fail at putting the proper metrics into the game, or fail to recognize its importance and skip it altogether.

If you’re not familiar with the term soft launching, it is releasing a game within a few territories with a limited marketing spend. It is a technique used by developers to gauge through analytics how well their game will perform before launching to major markets such as the US. Soft launching allows developers to gather data, spot bugs, and receive early feedback from players. Soft launching benefits almost all types of games. For games with PVP and multiplayer,  soft launching will provide you with the opportunity to test your server infrastructure and ensure it will be able to handle the load of a global launch. For free-to-play games and those with in-app purchases, it is important to determine how well the game will fare and if it has good retention and monetization mechanics. Simply put, this will tell you if your game will be profitable to launch or if you’d be better off moving on to your next project.

If you plan on looking for a publisher for your game, you should be aware that many publishers will look for some soft launch data before committing to any sort of monetary guarantee. Therefore, doing your own soft launch is highly recommended, as it will put you in a much better position during negotiations.

Before doing the soft launch, make sure to...

  1. Plan your soft launch and lay out your goals
  2. Make sure that the build is ready

Planning Your Soft Launch

Soft launches take time! While the median soft launch and optimization time for a game is 40 days, I have seen some games (including ours) take up to 6 months before it is ready for release. Be prepared! Establish the minimum or maximum number of days you would like to run the soft launch campaign. You should also think about how many versions you’re prepared to support during the soft launch.  Lastly, set the minimum number of players required to test your game.  From this, you can decide on your campaign budget or an alternate way to acquire those users (whether through cross promotion, PR campaigns, or whatever other methods you have). Do note that while the math here is straightforward, it can run up to a significant amount of cash.

CAMPAIGN BUDGET = Max Versions x Users per version x Cost per Install

Cost per install will vary from channel to channel (PR, Facebook Advertising, Ad Network) and country to country, so you need to do your research. There are a few ad networks that have published the average rates for user acquisition by country and genre. One such example is Chartboost which you can find at the following link - https://www.chartboost.com/insights. You also need a reasonable amount of users to gather data. In my experience, and this also depends on the game, I would say it takes around 200-300 users per version before the data starts becoming statistically significant.

Needless to say, this excludes the lion’s share of the budget which would be your development cost while soft launch is ongoing.

DEV COST = Team overhead x Months

It is important to know who your target user is so you can soft launch your game in a similar territory with behaviors similar to your target market. If you plan to release your game in an English-speaking market, you can opt to soft launch in Ireland and Netherlands, whose behavior is fairly close to US-based users. If you’re targeting the East Asian market (China, Japan, and Korea), a soft launch may be better managed through a local publisher. Though if you want a good mix of markets, Singapore and Hong Kong generally speak English but are more or less similar to China. If you’re not sure what your market is, then you have a problem. You can take a look at other games that are very similar to yours, and then check a service like App Annie or Priori Data to determine which countries they perform well in.

One trick we use is to do our soft launch in 2 phases. The first phase is to test out retention in a country with a much lower Cost-per-Install (CPI). For this purpose, we usually select a country or two in South East Asia such as Malaysia. Then as soon as you are happy with the retention results, expand to a country with a smaller market but with a high ARPU (Average Revenue Per User) to test out your monetization. Many developers typically use Canada or Australia if they are aiming for an English release.

Testing how well your game infrastructure can scale requires obtaining a large number of players. For this you do not need to worry about similarity in behaviour. Instead you can focus on cheaper markets. The Philippines and India have large groups of English speakers with avid players. Additionally, the Philippines and Indonesia have low acquisition costs and have an opposite time zone to European and American markets, allowing stress testing to be done outside the normal playing hours of one’s commercial market.  

Get your Build Ready

The primary purpose of soft launching is to track the game’s key performance indicators (KPIs). The success of a game after release is determined by how well it performs in terms of acquisition, retention, traction, and monetization. So before you soft launch, make sure that you have an analytics service integrated into your game and that you are sending the required events to measure all of the important metrics.  There are a host of third party tools like Flurry (free), Swrve and DeltaDNA. You also have the option to create your own but I would advise against reinventing the wheel.

Retention

You can’t make a game profitable unless players play your game for a long time.

To measure user retention, developers typically measure the percentage of users that return to the game on day 1, day 7 and day 30 relative to the date they installed the game; day 0 being the date of install. However, some developers will use the term day 2 instead of day 1 since they count day 0 as day 1 etc. It is also acceptable to use day 28 instead of day 30 since it is exactly 4 weeks from the install date and easier to plot.

What should your target retention rate be?? The higher the better right? Analytics companies say that an average game does around 30/15/7. I would say the minimum for a successful (and therefore profitable) game is 40/20/10. But of course, this varies depending on the genre and style of game. I would prefer a casual game to be at least 50/25/15, while a more niche but well made RPG or strategy would have a day 1 retention of only 25% but still be doing 15% on day 30 (not to mention much higher monetization rates).

Most games will not be hitting their ideal retention metrics on version 1.00!

So what to do? The first question your analytics should answer is what is causing your user to leave (and not come back). Analytics won’t be able to tell you if players are quitting because they do not like the game, but it should at least be able to pinpoint points of friction within your game that may be causing players to leave in frustration (or boredom). You should keep track of the people that open your app multiple times to try to figure out how many users drop out at certain points and stages within the game.  Also keep track of the onboarding funnel to find out if people “get it” and finish the tutorial. Then delve deeper, if they are quitting at level 9, maybe there is a bug or the level is too difficult. Find out which features are used the most and those that are  rarely used. Are there any ignored aspects in your game?

 

analyticsOur old analytics for a puzzle game with a Tableau Front-end, Flurry Analytics, and data stored on Google Big Query. We’ve since then shifted to a more robust analytics engine.

Hopefully, if your game visuals and gameplay is appealing (and more importantly, sticky enough) to your players, then you can focus on maximizing the retention in subsequent versions.

Remember that your goal here is not to let them start the game and let them play it to oblivion till they get tired of it (you won’t earn money or stay in the charts long this way), but get them to play in 5-7 minute chunks multiple times a day for the next 30 or more days. A great game can last years on the charts. This implies that you should:

  1. Have enough content on your soft launch build to last around a month (common estimates are around 40-60 hours of gameplay)
  2. Have a content pipeline that will allow you to have regular releases that will extend the lifetime of your game
  3. Have a way to ease out the player and make them stop playing, whether in an energy system or declining rewards for staying longer. There are a few free-to-play design blogs that tackle this issue that you can check out. Some favorites:

Monetization

Measuring the average revenue per user (ARPU) while soft launching in an alternative market allows you to see how your game might fare in your target territory. To measure monetization, pay attention to conversion rate, the amount of time it takes for a user to make a purchase, what users buy the most, and if there are particular user segments who spend more on purchases.

We are a VIP driven industry (that some of my industry friends call broken). A recent report by Swrve reported that only 1.9% of players purchased something during the month of February 2016 with 10% of the revenue from those players (0.19% of players) accounting for close to half (48%) of the total revenue in the mobile game industry. These are players that spend hundreds of dollars monthly and it has implications on your monetization strategy.

  1. Is there an upper limit to how much these players can spend? Some games may find that spending less than $50 will allow you to have everything you can buy in the game. Games at the top grossing charts have Average Revenue per Paying User (ARPPU) of more than $100.
  2. Is there a good curve? Many games employ some sort of diminishing returns where spending your first dollar gives a good boost (stats, abilities, progress, looks, etc), while the longer you stay in the game the more dollars you need to advance.

The trick is for games to give players emotional satisfaction with any purchase done in the game (whether it be $1 or $1000). Some developers may feel guilty with this approach. My opinion is that many other industries also have a ‘luxury’ approach, whether it be cars, hand bags, sports equipment or even airlines (a favorite presentation of mine from Kongregate).

clash royale

The geniuses at Supercell accomplished this elegantly by needing more and more cards to level up your deck. Either grind your way to the top or give in to IAP.

You also have to provide the player options. Some games rely on temporary boosters only, or just energy renewal, or cosmetic upgrades or permanent upgrades. The goal of a soft launch is to try out as many strategies as possible, and see which ones appeal to the player. You should look at the top grossing games, see what are they selling, and see if the same concept is applicable to your game. Finding a good combination of just the right amount of frustration (while still keeping it fun) and attractive packages (preferably permanent) to take out the frustration is the key to great monetization strategy.

Your goal here is to have as high of an ARPDAU (Average Revenue per Daily Active User) as possible to ensure that the players staying are spending on your game without driving them away.

Determining how soon players typically convert (to a paying player) in your game, as well as the specific event when that happens, allows you to tweak the game design to boost conversion. Also inspect which items, upgrades, or services most users are paying for, and which ones are being ignored. You may conduct split tests (there are SDKs for that) for the prices of items to find the most optimal price point. Lastly, get to know your game’s major payers. See if your game’s theme and design suits your paying demographic and if you should make adjustments to keep them happy.

How about Ads? Simply put, a game just relying on ads for the majority of their revenue will not generate enough cash to continuously advertise their game. While this is a good strategy for viral games that have a lot of organic downloads (think Flappy bird), relying on virality is very unpredictable and will more likely fail given the number of games out there. What is the chance that your game will be the one in 800 released on a daily basis to go viral? Some may argue that ads are a good way of monetizing the 98% of users who never spend any money on your game (i.e. watch videos for extra lives), but I’d prefer they keep playing your game and potentially convert to paying rather than having an advertisement send them to a another game.

Acquisition

There are multiple ways to acquire users for a soft-launch. It is much easier if you have an existing fan base or community that you can promote your game to (or cross promote from an existing game).

If you don’t have an existing fan base or community then you will have to do user acquisition yourself. However, this is much more challenging since it will involve some basic knowledge in digital marketing and advertising. You’ll also have to know how to create targeted advertising campaigns in Facebook, which will require you to make educated guesses on who exactly your target market is.

Since banners are relatively cheap to do, you can split test multiple versions of banners to get the best responses. You can track how well your acquisition funnel is performing by using analytics to measure the number of impressions, click throughs, and installs, and then use this data to optimize your campaigns and get the best possible cost per install (by testing multiple campaigns).

Also take note that monetization will vary greatly between countries, therefore cheaper installs does not necessarily mean you will get higher profits. Another alternative is to do user acquisition via ad networks by paying for installs. However most of them will charge a fixed rate per install and in many cases will be more expensive than running your own advertising campaigns.

Organic Traction

Organic traction is the combination of many factors, such as how easy it is for users to find the app, how many users recommend the game to others, and what the general feedback is regarding the app.

Is your game visible or easily searchable in the app store? You should check whether your app description and pictures are compelling enough for potential players. Before soft launching, you will want to have social features integrated into your game. This will allow you to measure how many users are freely recommending your game to their friends or how effective your reward system is for referrals. Most importantly, take note of the ratings and comments from users on both the app store and your social networking pages. Analytics can tell you if there are areas that need work, but it won’t exactly pinpoint what it is and how to solve it. Users will usually report any bugs you missed during initial testing and let you know if you need to make any changes to your mechanics.

Ultimately, your goal is to have players invite others to play the game - whether through facebook, other social networks, messaging centers (Line, Kakao) or word of mouth. If each player you get through user acquisition invites one more friend, then effectively, your $1 CPI becomes $0.5. Of course, being on the charts or featured by Apple or Google will help immensely, but relying on only these will not make your game sustainable.

Putting it all together

The goal here is to have the Total Average Lifetime Value (LTV) of your game surpass the Acquisition Cost (CPI). The magic formula being:

LTV > CPI = Winning game

While there are a few mathematical models to project your LTV using some initial metrics and data (Retention and ARPDAU), nothing beats testing it out for yourself to see what the average revenue of a user is over time and how long (if at all) it takes to recoup the cost of acquiring users. You can segregate this per version, per campaign and per territory. Determining how soon your game breaks even (if at all) is the ultimate goal of your soft launch. Ninety days is a good goal, although the sooner the better and some games may take just weeks or days before it breaks even, unfortunately, the vast majority will not break even.

To Iterate or not to iterate

What should you do if the game is not reaching its metrics or breaking even after all the months or years of hard work you put into developing it? In this scenario you’ll have to make the decision. Either iterate or cut your losses and move on.

Is the game’s retention or monetization close enough and can it still be improved? Are there some major bugs or missing features that can improve its metrics? Is there a flaw in the fundamental design of the game that doesn’t resonate well with people or cause it not to monetize?

This is a judgement call you will have to be prepared to make. This is a tough, often heart-breaking decision as I myself have been in projects that have not been launched. At Playlab, we’ve developed 10 games, soft launched 8 of them and just released (and continue to maintain) 3 of these games. Some games will simply fail to make the cut because no matter how awesome we think the plan and the game is, they just fail to hit the mark.

Wrapping Up

The importance of doing a proper soft launch is often overlooked by developers. But taking the time to measure your KPIs and analyze your data may determine whether your game makes it to the top or simply fades into oblivion after a global release. Soft launching gives you feedback if the game is ready or not. It’s best not to waste time and money by launching a game that won’t recoup its investment.  

Looking for a partner in publishing your fun, free-to-play, casual/mid-core game?

Submit it to us at Playlab:


Related Jobs

Game Circus
Game Circus — Addison, Texas, United States
[09.26.17]

Backend Web API Engineer
Insomniac Games
Insomniac Games — Burbank, California, United States
[09.25.17]

User Experience Researcher
Insomniac Games
Insomniac Games — Burbank, California, United States
[09.25.17]

FX Artist
Deck Nine Games
Deck Nine Games — Westminster, Colorado, United States
[09.25.17]

Mobile Programmer





Loading Comments

loader image