GamesBeat

A great app or game is required, but not sufficient, for success in mobile marketing

Epic Empire

Above: Epic Empire

Image Credit: Pocket Gems

Michael Dawson is the vice president of business development at Pocket Gems.

Very few of the 1.7 million apps and games on the Apple iTunes App Store and Google Play are truly successful. We believe that’s because few companies know how to marry great product and great marketing savvy. Let’s be clear: You need a high quality app, and nothing less. But a great game alone usually doesn’t suffice.

Secret Passages

Above: Secret Passages

Image Credit: Pocket Gems

Every now and then, a new app or game will get prominently featured or will go viral without much effort on the developer’s part. Unfortunately, most apps don’t go viral. How, then, does one even begin to chart a path to give it a standing chance to be discovered and enjoyed by millions of people?

We have found that the very first step is to look at who you already have a relationship with; you should always start by reaching out to your existing customers who already love your product(s) if you already have a player base. It’s a pretty significant step one for us: More than 50 percent of Pocket Gems players have played more than one of our games. To be fair, we have heavily invested in creating advanced methods of promoting our new games to our own loyal player base over the past four years, it did not happen overnight. Nonetheless — any developer that has a game in the market has a starting point.

The next step, as every app maker knows, is the difficult and expensive: paid marketing. In fact, given the lack of organic discovery opportunities such as networks like WeChat, Line, and Kakao in North America and Europe, paid mobile marketing is more critical than in other global territories. The majority of top-grossing apps, besides being incredibly high-quality products, buy a meaningful number of ad impressions each day and hope that their average revenue per customer exceeds the average spend per customer.

The bad news is that finding customers who will love your app, profitably, is incredibly hard. The perhaps good news is that there a lot of expertise out there: We have poured four years into mastering the art and science of paid mobile marketing.

We learned a lot — some through successes, and some through mistakes. Let me share five areas that we had to wrestle with to illustrate the amount of sophistication required to deploy mobile marketing effectively. My hope is that by sharing insights we have gathered in the process of building Pocket Gems’ distribution function, where we’ve amassed over 150 million installs in four years, you will get there faster.

1) Deploying too many channels at once, and not reassessing your chosen channels

It’s impossible to test a large number of channels at the same time, so it’s best to start simply, prioritize, and reassess because appetites, the marketplace, and a channel’s inventory will often change. We used to test any partner that we came across (and tested over 148), but now we have a screening process for new channels before starting a test. We now have 57 ad networks and marketing partners that we work with at any point in time. But the list and approach keeps changing. Facebook was not a partner just one year ago, and now it is one of our largest channels. And we used to buy more banner ads, whereas today we always prefer providers who can deliver a high volume of full-screen interstitial ads. There are a lot of potential partners for performance advertising on mobile; you will need to navigate, prioritize, and reassess over time.

Mobile marketing landscape

Above: The mobile marketing landscape.

Image Credit: Pocket Gems

2) Creative assets that are fixed and forgotten

Advertising design can get tired and stale fast. Ads and messages need to remain fresh in an environment that is literally getting bombarded with new products. We learned that it’s best to be ready to invest in new creatives that show off your great app up front. Often, you get attached to that perfect creative that you labored over. But to attract the greatest number of people, you need to regularly refresh ads every week or two, and you need to test those refreshes to make sure people agree that you are making good decision. The more regular and systematic you can be about this, the better the results will be.

3) Acting as if all players are created equal and that all installs should have the same price

We learned that looking at the average cost-per-install (CPI), or what you pay to get one app installed, is a gross oversimplification of what the values are. Early on, we noticed that you need to at least slice by geography and device type on each ad network. For example, players on an iPod Touch in Brazil will, on average, spend dramatically less than iPhone 5S or Samsung S4 players in Japan, so you should bid differently for those impressions. Geography and device are great initial buckets to analyze. Over time, if you don’t have over 100 bids across your various networks, there is probably room to improve. Check out this great CPI map by one of our partners, Chartboost.

Paradise Cove results

Above: Paradise Cove results.

Image Credit: Pocket Gems

4) Underestimating the complexity of tracking where customers come from

Tracking customers is complicated and can be downright aggravating. To know which customer clicked on which ad before installing your app is not easy. And if you do not know where a customer came from, you will never be able to figure out which of your channels are working and which are not. This happens because tech can break, either on your side or from your ad network. And given there are different methods of tracking, redundant and overlapping results can occur. Solutions like HasOffers help you avoid building out a full server-to-server tracking infrastructure like we have, but they are still not 100 percent accurate. It might be advisable, especially if you’re just starting out, to just limit yourself to a few channels so you can manage the complexity and track your players effectively.

5) Not knowing if you are marketing profitably

If you don’t know what your margins are, it’s impossible to manage your business well. It’s critical to measure profitability for each bid on each channel in order to understand the efficacy and profitability of your efforts. Otherwise, the average ROI could be hiding money-losing bids, and could be masking areas where it makes sense to spend more.

The theory is easy — you want to be sure that you are making money for each bid you have in the market. But this simple idea is extremely hard to execute: If you have tracked where each customer came from, and you know how much you paid, you can then match that data up with how much revenue those customers are generating (removing outliers) to calculate ROI. For Pocket Gems, after an arduous process, we have built a system to automate this and brought our reporting time down from a week to two hours. But no matter how long it takes, you need to know bid by bid if you are making the right choices for your business. You will then want to project this over the lifetime of that particular customer. In other words, you have to find a way to predict that lifetime revenue with limited data. From there, you can decide if you should keep that bid as is or change it.

Michael Dawson of Pocket Gems.

Above: Michael Dawson of Pocket Gems.

Image Credit: Pocket Gems

In summary

You have options when it comes to managing your mobile marketing. If you don’t have the resources, the depth of experience, the talent and the bandwidth, you can always partner with a publisher or another third-party that can manage your spend. Whatever you decide, make sure you or your partner follows the rigor described here. It’s a big task but ensures the great app you’ve invested so much time to build stands out among the 1.7 million apps competing for attention.

Michael Dawson is the vice president of business development at Pocket Gems. He has driven many of the company’s user-acquisition partnerships and platform relations. He wrote this guest post for GamesBeat. 


VentureBeat is studying mobile marketing automation. Chime in, and we’ll share the data.