After capturing a click, the Web_Click_Through_Rate model is showing a response, but it is not captured as either positive or negative. Is this expected?
The click was executed and the number of responses incremented to 1 at about 4:45 PM, and after checking an hour later, the response still isn’t categorized as positive or negative.
The response timeout in the Predict Web Propensity prediction is set to 30 minutes.
See attached screen shot of the Adaptive Model Management Landing Page.
That was the first thing I checked. pyOutcome is recorded as “Click” in our system. I updated the adaptive model outcomes to look for “Click” instead of “Clicked”. Prior to that update, the models weren’t capturing any responses, just recording default (negative) responses. Now they are capturing responses, but I was also expecting the number of positive responses to increase with each click. In the screen shot you can see total responses = 1, but both positive and negative responses = 0. My interpretation of this is that the model has recognized the click, and because of that, it’s not recording a default response, but it’s also not recording the click as positive for some reason. Is there another place that I need to update “Click” vs. “Clicked”?
Unfortunately the attached screenshot isn’t opening for me.
The update frequency of an adaptive model is determined in the settings of the adaptive model rule under ‘Update model after every .. responses’
An update can also be manually triggered in the Model Management landing page by selecting the model and clicking update, that should result in an increase of #Positives/#Negatives/Processed responses (after refreshing the screen).
If that doesn’t happen when you click update, please create a support request