Blog Series: The Status of View-Through Attribution (Part 2)
👋 Welcome back to our mini blog series on View-Through Attribution! If you missed part one (our interview with marketing analytics platform Singular), you can find the blog post here.
1. What percentage of the video ad spend that you see, would you say is VTA enabled?
Approximately one-third of our clients are using impression signals (some for attribution).
2. What is the typical VTA window? What’s your perspective on it?
The typical window is between 2 to 24 hours.
There are two ways to think about View-through Attribution. One is as a historic exercise with incrementality in mind and the second is as an analysis of direct responses that occurred as a result of an ad unit served. The goal is to determine whether a user installed the app organically or was influenced by an impression. No ad unit or interaction is the same, and each should be considered uniquely.
With sufficient data, we can look at overarching trends, such as mean-time-to-install distribution and user quality over time. These data sets allow marketers to make statistically significant judgement on what lookback windows are most relevant to any given campaign, creative, or publisher. Seeing older impressions or disqualified impressions is also valuable using our lookback windows or Query tool.
3. What are some of the benefits that advertisers have seen when enabling VTA (please share real numbers if possible)?
Advertisers often see up to a 30% decrease in “organic” traffic and gain deeper insight into the value of their placements.
4. What are the main reasons why some advertisers are still skeptical?
With VTA, marketers often think they are paying for organic users; however, it’s primary intent is to determine lift. Marketers should be wary of VTA if their attribution partner matches impressions to conversions regardless of the time frame. Their conversion rate might be higher, but the data is poor. Proper attribution partners do everything they can to mitigate these incorrect attributes.
5. As an industry, how can we generate more trust in this attribution system?
As an industry, we can generate more trust in attribution by helping marketers obtain accurate data. As a mobile measurement provider, we’ve made it our job to operate as an unbiased referee that helps reconcile data to provide accurate attribution. We can only do that with the highest integrity signal from our inventory sources. Only then can we assist marketers in cleaning up their ad signal to have separate impression and click data streams. Enabling anti-fraud tools further cleans the signal along with tight lookback windows (down to minutes and hours as opposed to days). For example, the accuracy of fingerprint (probabilistic) attribution is 98% within the first 10 minutes. That accuracy drops drastically to only 20% after 7 days. With Kochava, marketers can adjust lookback settings for probabilistic clicks, deterministic clicks, probabilistic impressions, and deterministic impressions by tracker and down to the minute.
We will be back next week with more third-party opinions about View-Through Attribution and a blog post on the topic from Chartboost’s Chief Strategy Officer, Pepe Agell. Stay tuned! 👍