- Ad mediationUnity LevelPlay
- In-app ad networkironSource Ads
- ironSource ExchangeironSource Exchange
- On-device solutionsAura
- Player ownership solutionsAstra
- App marketingLuna
- Ad mediation
- The basics
- The best practices
- Featured content
- Log in
- Sign up
Choosing the right attribution partner
Next in this course:
In this last class with Lara Doel from Singular, we’ll learn how to make attribution a seamless part of our user acquisition strategy. Lara shares 3 qualities every attribution platform should offer you, so be sure to take note.
Now that you know what Attribution is and why you need it, the next point of discussion is how to effectively make attribution a seamless part of your marketing activity.
With more sources, more data, and more tools than ever before, the modern data explosion has made marketing data anything but simple. Most marketers are not data engineers by trade, so they need an attribution and analytics provider that allows them to ditch the spreadsheets, automate data collection and processing, and get straight to uncovering performance insights, which will help them optimize their campaigns and reach their goals.
As a marketer, you would choose an MMP vendor to work with and then install an
SDK in your app, which is a line of code from the vendor, allowing you to track events and activity. That vendor will then do the work for you of collecting the relevant in-app data and presenting it to you for analysis. This is the first part of measuring the outcomes, and this data needs to tie back to the marketing investments.
If you’re stuck trying to aggregate campaign data from multiple ad networks and combining it with conversions and events, all in Excel, you’re not only drowning in tedious data transformation tasks, but your conclusions will inevitably be inaccurate.
No matter how analytical a marketer is, manual data collection leaves room for human error and data that doesn’t match up on a multitude of levels. As an example, every ad network or publisher structures its campaign reporting and classification differently. There is no standard across them, making it difficult to aggregate and normalize campaign data to get a side-by-side view of performance across ad partners.
To summarize, you need the help of a company that is focused on collecting and cleaning the data in order to hand it back to you in a format that you can digest.
As a way to create a better, more organized system across the board, almost a decade ago, Facebook decided that app install campaigns needed a proper system for third-party independent measurement. They called those third parties mobile measurement platforms, known as MMPs. Today MMP is the generic name for mobile measurement companies that verify marketing impact on literally thousands of platforms and ad networks at a time.
Can you do it alone and build your own infrastructure? Sure, but it will take in the region of 500,000 working hours to get that up and running, and then you need to take into account the ongoing maintenance, in-house expert team that you must have available, and the cost of using those resources on a project like this, as well as understanding what projects will have to be pushed off with this on the table, and of course, just basic time to market…is it really worth it??
For the purposes of this course, I’m going to detail the how of working with an official MMP with 3 key points that you need for effective attribution.
If we’re honest as marketers, we love the creative stuff. We love ideating and implementing and promoting. But to prove the effect and impact of our work, we have to rely on the numbers. But, to do our jobs well, we need the ability to collect and analyze those numbers in the most efficient way to save our teams the time to be creative.
If all user journeys were the same, all apps could use the same exact reporting. But this isn’t the case; each app is unique. And the KPIs you need to hit are too. So your reporting tool need to be customized to what acquisition, monetization, and retention means to you, and ultimately report on the KPIs that are tailored to your business.
Even more challenging is trying to unify this upper-funnel campaign data with lower-funnel attribution data. Campaign data is reported in aggregate, while conversion data (except for SKAdNetwork conversions) is reported on the user-level. It needs to be standardized, merged, and enriched to provide actionable insights.
You need to be able to see a snapshot of all your data at once, across all your apps, and then drill down into deeper granularity with a single click … and further customize that drill-down by any dimensions that you choose.
In order to measure your entire user lifecycle from acquisition to re-engagement and retention across all platforms (think mobile ad to app or even mobile ad to website to app), you need flexible attribution models – with all the data automatically leading back to a structured report. We’re not going to go into all the attribution models now, that will need to be for another course.
Your MMP should be able to use the SDK that you implement to collect, organize, and present the data from all your marketing campaigns, across platforms and devices, to provide brands with a unified view of your campaign performance. In other words: a single source of truth for you to rely on, relate to, and analyze from.
Another key element to this is to ensure that your data is protected against fraudulent activity with the right detection, prevention and rejection methods to ensure that the your reporting is correct
100% integration capabilities
To make this possible, your provider needs to be certified to work with, and integrate with, every advertising source and network that you work with so that you’re receiving a complete view.
Your data where you need it
If you work with an internal database, it may make sense to have your data available for analysis from there. A well-built ETL, meaning a system to Extract, Transform and Load data, eliminates the need for you to build data pipelines, plan schemas, handle errors, monitor and scale your pipeline, and handle late arrival data. This saves precious engineering resources and ensures that your data is extracted, organized, and transformed without you even needing to think about it.
The best practice is to choose an MMP that provides you with
1) Extensive reporting
2) 100% network coverage and
3) Seamlessly extracts, organizes, and transfers your data to wherever you need it
Ok everyone, we’ve come to the end of our course on the what, why, and how of attribution. This topic is a big one. There are so many elements that go beyond what I have detailed in these three courses but I hope this has given you a basis to start your journey into Attribution with the right modeling and mindset.
Thank you for joining me and good luck!!