Tracking the value of mobile app advertising
In the beginning there was the Web, and the Web was good, but for online businesses to thrive we needed advertising.
We have come a long way from the first unclickable Internet ads for Sears in the 1980s to a $26 billion online advertising industry, yet you have to stop and wonder how we have come so far, so fast.
I suggest that tracking technology played a crucial role in the value of online advertising by providing real, quantifiable results for advertisers for the first time.
Unlike billboards or Nielsen ratings, advertisers can actually study how many users saw their ads, clicked their ads, and even continued on to make a purchase, and that is just the most basic information available to advertisers.
This ability to track user behavior birthed an entirely new industry of analytics and tracking technologies for analyzing and optimizing online advertising efforts.
Why the history lesson?
How the cookie crumbles
For how far we have come in tracking and optimizing online advertising, we are still in the dark ages of mobile advertising, especially when it comes to promoting mobile applications.
I feel like I stepped into a DeLorean and ended up back in 1985, with no way to see which advertising is actually bringing value.
Sure, there are already major ad networks and solid traffic sources for promoting mobile apps, but how do you know which advertising and marketing channels are bringing you new users or app installs?
Beyond that, how do you know which advertising efforts are bringing you valuable users who actually spend time on your app, progress to higher levels or purchase products?
Think you can just plug in your marketing analytics platform you use for your Web site? Think again.
Tracking mobile apps installs and other events back to a user’s click or marketing source actually poses some unique problems.
For one, the most common way to track online advertising today does not work on mobile.
Even if you do not do any online advertising yourself, you have heard of cookie tracking.
The basic idea is that a “cookie” is placed on a user’s browser when she clicks a tracking link. When the user lands on a conversion page a tracking code checks the cookie to know where the user came from.
We have been using cookie tracking on the Web since 1995, and it continues to be the most popular tracking method online.
The problem with mobile devices is that they treat browsers very differently and, in addition to that, app stores or marketplaces do not use browsers at all.
This is further complicated by the fact that mobile apps themselves do not operate within a browser. So, to track mobile app installs back to a click or impression, you will need a different solution.
If you are clever, your response to this is, “I’ll just use server-to-server tracking then.” No, you will not because that does not work either.
When a user enters the iPhone app store for example, you cannot pass in any unique information that you can collect from the user on the other side.
In other words, the iPhone app store is like a cold dark place where Web tracking goes to die.
I know it is hard to believe, but there is no clear way to optimize advertising and marketing for mobile apps.
Studies from major mobile ad networks such as InMobi, Flurry and Fiksu all point to mobile advertising growing at an alarming rate.
In the case of InMobi, the company reported that its numbers grew by 50 percent in 90 days. And still, advertisers and app developers are not tracking their own campaigns back to installs and in-app events.
For most, the only metrics they are working with are how much they are spending and how many new users they are acquiring over all. If you told a Web marketer to run his campaigns blind like this, he would call you crazy.
I can honestly say I was astonished when we first stumbled on the lack of tracking for mobile advertising, and the more I learned, the more I was amazed.
You mean there are massive ad networks that are growing explosively and advertisers do not even know which installs come from them or from their own viral or organic marketing efforts, much less which customers came from which ad networks?
About six months ago, one of our larger customers in social games asked if we could figure out a way to track mobile app installs back to their source so it could work directly with publishers.
The customer had all these great mobile publishers and was working with top ad networks, but it needed to know which installs were coming from which efforts.
“No problem,” we said, “should be simple enough.” Wrong.
Our immediate solution was to use server response tracking like I mentioned before and then to just use IP address, but the accuracy was terrible. That made us take a closer look at the problem.
Some of the ad networks that are tracking app installs largely depend on device ID or what many call UDID (unique device identifier).
Until recently, this was a very accurate way to track individual users, but Apple recently announced that it is phasing out this unique identifier.
This caused some ad networks to hop onto tracking a user’s Mac address, but, of course, Apple will shut down this identifier as well.
New game in town
There has to be a tracking solution with a high level of accuracy that does not depend on these personally identifiable IDs.
So far, we are finding that mobile app game producers are the most knowledgeable and aware of the tracking problem. These are the kinds of businesses that pump out new games almost on a weekly basis.
Companies such as Zynga, Big Fish Games, GameHouse and PopCap are some of the most mature companies in mobile apps, and they really get it. They have already been working with ad networks, aggregators and other mobile advertising solutions.
Some of them even promote their own apps through other apps they own but still have no way to track which installs came from where. These are the companies that really understand the need for non-biased tracking.
At the end of the day, the greatest problem with tracking mobile apps installs, and what makes it so different from tracking traditional Web advertising, is the Apple and Android platforms.
In this case, Apple and Google’s Android really hold all of the power.
No matter how compliant advertisers are with tracking and analytics, Apple ultimately controls what code it will permit on an Apple device, and though Android is far more relaxed, the same is true.
It will certainly be interesting to watch how Apple and Google continue to exercise that control and where their master plans will take us. Giving advertisers eyes to see their advertising campaigns is crucial to the growth of their platforms.