Receive Alerts on Key Metric Changes
February 20th, 2017 Ryan Floersch
With so much data to continuously explore and keep track of, it is important for your programmatic analytics to proactively alert you of the important things that need your attention. With that goal in mind, we have enhanced our alerts feature to now include percentage change conditions.
Percentage change alerts give you the ability to proactively monitor large increases or decreases in key metrics that you actively track on an hourly or daily basis. For example, you can now receive alerts if:
- Total spend increases by 15% or more compared to yesterday
- Click-through rate for a high-touch campaign has decreased by 30% as compared to the same day, previous week
- Delivery rate of impressions has decreased by 40% over the course of an hour
By leveraging percentage change alerts, you can spend more time investigating the causes of large increases or decreases in key metrics, instead of manually identifying these trends from your data on a recurring basis. In our example above, a large increase in spend would be a good indicator that you should investigate if any new campaigns have started that might need your attention to help optimize. Or, if delivery rates had dropped dramatically over an hour, you would likely want to quickly identify if any integrations are down that may have caused the sudden decrease.
How Does It Work?
Percentage change alerts are now included in the alerts feature found on your Explore Dashboard. Click on the bell icon located on the top right of your screen and an “add an alert” pop-up window will appear. First, give your alert a customized name. In this example let’s use a percentage change alert to track “Spend Decrease,” as I want to be notified if overall spend drops by 15% or more on any given day.
Under the conditions section, select “Total Spend” from the metrics dropdown, “decreases by %” from the condition dropdown, and enter “15” in the percentage field. In the period fields, we will want to “check every day” in comparison to the “previous day.”
Since alerts have the potential to become overwhelming, you will notice at the bottom of the pop-up window a frequency estimate. Based on your data history, this estimate lets you know how many times you would have been alerted over the course of the last 30 days. To decrease the number of estimated alerts you might receive, we recommended adjusting the percentage field until you get to an estimated volume of alerts that make sense for what you’re trying to accomplish.