Signal Clarity is the Foundation of Campaign Data

Race car drivers know optimal engine performance depends on the quality of oil they use. Mobile attribution and measurement work similarly based on the quality of data signal sent and received. The richer and cleaner the signal sent by media partners to the Kochava attribution engine—the more accurate the attribution and ability to analyze trends, and optimize marketing channels. Signal clarity is an assessment of the integrity, reliability, and breadth of the data stream flowing between a media partner and a mobile measurement provider (MMP). Every part of the signal is important for optimization, from creative IDs to in-app events.

Rich data from media partners to MMPs

For the attribution engine to work best, it needs good campaign data from partners so the MMP can attribute and in turn provide solid data for measurement and analytics. Data sent from media partners to an MMP come in the form of impression and/or click records that describe the serving of an ad (impression) and a user’s engagement with that ad (click). Within these impression and click records, partners can pass valuable data elements which facilitate optimization and deeper attribution insights. The key elements that contribute to a good signal include:

• Impression stream: device ID, site ID, creative ID
• Click stream: device ID, site ID, click ID, creative ID

Having these elements passed in the impression and click streams allow marketers to perform a number of tasks to optimize their marketing. Knowing the device IDs reached in a campaign leads to more accurate attribution. Knowing creative IDs allow marketers to perform A/B testing to compare and hone their messaging. And site IDs inform them of how well the partner’s channels deliver ads. These elements not only inform marketers but also enable publisher optimization. In addition to knowing site IDs, signal clarity also considers its life cycle. The life cycle of a partner’s site IDs is another meaningful factor in determining the overall strength of the signal. This aspect of the metric considers a partner’s percent of low cardinality sites. The translation—partners should have a reasonable volume of clicks across their sites—not thousands with short life cycles that elicit one click on each. Such minimal traffic on short-lived sites is meaningless. Not only do these sites create doubt in terms of legitimacy, but they also leave the marketer with little recourse for publisher optimization strategies. Questions to ask your media partner about the data sent to your MMP:

• • Are they sending clicks only, or impressions and clicks?

Sending separate data streams should be an innate capability, but it’s important to ensure that your partner is sending the two individually.

• • Do they report site IDs to help optimize on the publishers that are driving the best traffic?

Site IDs allow marketers to employ view-through attribution and click-through rates useful in analyzing certain types of media such as video.

• • Can they pass creative IDs to facilitate efficient A/B testing?

The ability to A/B test is invaluable. By knowing the creative IDs used, you can modify your messaging for future campaigns.

Rich data from MMPs to media partners

Data sent from MMPs to media partners are postbacks that report installs and/or in-app events that occurred as a result of the ads served by the media partner. These give partners real-time feedback for campaign optimization, machine learning, partner reporting, and other functions. Not all media partners can receive the same type of postbacks, and it’s important to know what data they are capable of sending to and receiving from MMPs. Questions to ask your media partners about data they receive:

• Can they receive installs and in-app events?
• Installs are an indicator of successful acquisition, but knowing which users converted through in-app events tells a broader story of how valuable those acquired users are. In-app events inform marketers on CPA (cost per action) models based on campaign goals and also perform auto-optimization (app event optimization) of certain events based on inventory.
• Can they observe a percentage of users delivered by a particular segment of inventory from a publisher?
• Marketers can see which users completed an event and use the information to allocate ad spend to a particular channel. They can also apply the information to suppress certain audiences for reengagement of certain users.

These questions and more form a basis in assessing a partner’s signal clarity.

How to clean up and protect your signal

To start, work with partners to ensure that they are sending click and impression data streams separately. Push for the inclusion of site IDs, which provide valuable insights of user quality and other metrics by the publisher. Implement anti-fraud tools to catch errors/fraud such as impressions sent as clicks or site IDs with high click volume. To compare media partners performing well in the metrics of signal clarity, fraud mitigation, quality traffic, and correlation, download the latest version of the Kochava Traffic Index. In the report, you’ll see top-performing partners, such as Leadbolt. The report will be released quarterly so you can observe which partners maintain or improve their ranking. “We understand the value of the data points necessary for a strong signal and what they contribute to optimization. It’s important for us to work with clients to ensure we’re supplying what they need, and it’s part of our longtime commitment to delivering quality users globally at scale,” said Dale Carr, CEO and founder of Leadbolt. Signal clarity is the foundation for achieving accurate measurement and is the basis of future marketing decisions. If, as a marketer, you can’t trust your signal, you can’t trust your attribution. And, if you can’t trust your attribution, you can’t trust your measurement, traffic quality, or the validity of campaign data. For more information about signal clarity and the metrics that contribute to more accurate attribution and measurement, download the Kochava Traffic Index.