![tatari-lockup-black.png]](https://tatari-5147324.hs-sites.com/hs-fs/hubfs/tatari-lockup-black.png?height=40&name=tatari-lockup-black.png){kind=small}
What is a category DragFactor? How does Tatari validate that?
Category DragFactor is the average of all advertisers within the same industry. As of March 15, 2021, all new Tatari clients start with a category DragFactor until we’ve accrued enough data to calculate one that is customized to your business.
Custom DragFactors tend to cluster together in the same range based on what industry they are in. Historical data analysis shows that the industry average is highly correlated with custom drag factors, and thus this is a good proxy in the absence of a custom DragFactor.
Once a custom DragFactor is established, how often will it be updated?
Custom DragFactors are refreshed monthly (subject to having sufficient data). Reach out to your Client Services Manager to review the output and decide when to implement it into your reporting.
I currently have a default DragFactor. Why is 2.2 the default?
For Tatari clients starting prior to March 15, 2021, the default DragFactor was set to 2.2 - a conservative estimate as this was the lowest DragFactor computed across all of our clients. As our portfolio of clients has expanded, new clients as of March 15, 2021 will have a category DragFactor, which is the average of similar brands in your industry.
Why do we use a DragFactor on streaming if we have 100% IP matching?
DragFactors are still employed on streaming so that we can provide you next-day results by translating immediate response to total response.
Does the DragFactor assume delayed response converts at the same rate as immediate response?
In short, DragFactor is solely a metric to measure total response, not conversions. A response is when a user takes action upon viewing an ad, such as a website visit or app install. A conversion is when a user moves from one stage of the funnel to the next, such as when a user browses your site and eventually purchases a product. We define it as total response divided by immediate lift.
Note: We almost always find that delayed responders typically have considerably higher conversion rates than immediate responders.
Is a control group removed from the Drag Factor calculation? If so, how exactly is that created?
Yes, we first compute the view-through response for those exposed to an ad. We then subsequently subtract out a control group (composed of similar users) to eliminate the response for non-exposed users. Subtracting these distributions results in the incremental response. From this, the DragFactor is calculated as the final response vs. the immediate response.