Moving from app cost per install to profit per install
By Dan Kimball
You are a mobile applications marketer. You spend $100,000 on 200,000 installs for your app. Fifty cents install. Not bad, considering the market average is just over $1, right?
Well, it is certainly not a sure thing for your business’ ROI if you are a free-to-play app that relies on in-app purchases as your monetization strategy. The concept of free-to-play describes any type of app that is free to install, including mobile commerce and other types of apps.
The bottom line: Not all installs are created equally, so how do you pinpoint which installs are valuable to your business and move the needle on revenue?
I have been a digital marketer for most of my 15-year career, starting with buying banner ads on Web sites for clients while working at an agency as a media buyer.
Campaigns were managed with great analytic rigor. Regardless of how we purchased the media – cost per thousand, cost per click or cost per acquisition – we built and optimized our campaigns through this four-step process:
1. Buy minimally-viable amount of media to provide enough sample for site-level and creative-level analysis
2. Analyze campaign, and hand-pick which cells – sites and creative – were most “profitable”
3. Invest heavily in the sites and creative that were most “profitable,” and scale user acquisition efforts
4. Rinse and repeat across each marketing channel
Digital marketers are always looking for ways to make the smartest media buying decisions.
Tools such as DoubleClick DART and Atlast DMT have enabled us to run incredibly dynamic media buys, optimizing in near-real-time against CPC and CPA benchmarks.
Google AdWords revolutionized media buying by introducing machine-learning techniques that removes human intervention from the process altogether.
Now, in the new world of mobile app marketing, marketers can buy installs on a cost per install (CPI) basis. And some ad providers are taking it a step further, introducing “cost per engagement” campaigns.
This type of product minimizes the advertiser’s risk even further by reducing the upfront investment in potentially low-quality installs, and putting more of the risk in the hands of the ad networks.
But is that enough to guarantee a worthwhile return on ad spend?
Look back at Nos. 2 and 3 in the aforementioned media campaign development process. And, look more specifically at the word in quotes: “profitable.” Were we really measuring and optimizing around the profitability of the ad campaigns? Not really.
We were making the most intelligent decisions we could about where to invest our clients’ money based on the data we were able to access at the time.
Campaign effectiveness was determined by analyzing which cell of the campaign produced the greatest initial cost per action (CPA). A single action.
Success was defined, measured and optimized on the very first action completed. Person sees ad. Person clicks ad. Person completes action (install, purchase, registration, etc.). Campaign is optimized based on which channels and which creative produce the most first actions for the lowest cost. Rinse and repeat.
Now, in the mobile app free-to-play model, it is incredibly challenging to make this method of measurement and optimization work for your business.
Since installs are free, most app developers are managing their campaigns to the lowest effective CPI, and then conducting some analysis post-hoc to determine if the installs were high quality or not.
Ultimately, this not a dynamic method for optimizing ad spend when the ROI inputs are far from real-time.
But what if we could evolve our perspective on ROI? What if we could frame our measurement and optimization around a much longer-term outlook on the value of the installs we are buying, well beyond the initial install, or even the initial engagement within the app?
And what if that data were available to us quickly enough to make game time decisions on marketing investments?
Whereas a digital marketer’s campaign-level ROI has been defined for the past 15 years by the following equation:
Initial Transaction Revenue – cost to acquire a customer (CAC)
The new and improved way of measuring campaign-level ROI is:
Customer Lifetime Value (cLTV) – CAC
With today’s analytics tools, mobile marketers have the ability to measure and optimize their marketing programs well beyond the initial customer acquisition “return.” They can now analyze how well each of their installs perform, and attribute each back to the original source of acquisition, thereby understanding which media buys are yielding the greatest quality.
Most importantly – and for the first time ever – marketers can measure and optimize their campaigns based on the predictive lifetime value of their campaigns, right down to the creative that generated the initial click.
THIS NEW METHOD of measuring and optimizing based on the cLTV of my ad buys is now truly using “profitability” as the KPI. It is no longer about CPI (cost per first action), but rather PPI (profit per Install).
Making this transition from CPI to PPI is a longer-term outlook on the value of customer acquisition channels. It is a redefinition of what we mean by “return” in the ROI equation. And, at the end of the day, it is just smarter business.