Posted: 9 Dec 2020 14:51 EST Last activity: 15 Dec 2020 20:55 EST
Function of Contact Policy Lookback Window
I need some clarification to provide guidance to a customer on how to define the "lookback" window for contact policies. By lookback window, I mean the first part of the contact policy that states, "If there are X (outcome)s in the last Y days...", "Y days" is the lookback window.
Can you explain why contact policies are constructed with a lookback window? Why not just a suppression period? In my conversations with customers, they have tended to verbalize suppression rules in terms of a trigger event and a suppression period, without referencing a lookback window.
Here are a couple examples.
If a customer is sent an Acquisition action by email, a business may not want to send any more Acquisition messages by email for 7 days, to give the customer a chance to respond. The 7-day suppression period should start when the email is sent, rather than when it is opened (when a response would be captured). Additionally, a lookback period of 7 or 30 days shouldn't be necessary in this case.
As another example, a business may want to display a monthly Service message in a self-service portal, and when a customer clicks on it, suppress that action for 30 days in the channel. The 30-day suppression period should start at the time of click. If the suppression logic is run at the time the click is captured, then why is there a need for a lookback period for this use case?
There may be other use cases that I'm not thinking of where a lookback period is needed.
One more question. Where is the lookback window used in the strategy logic? I didn't see it used in the BehavioralLimits strategy or its substrategies (though I could have missed it). I don't think it would be used in the NBA Strategy Framework, since contact policies are precomputed.
When defining contact policies you can set the triggering outcome to be whatever you like, so for your email example you would set the outcome to be Pending, which would track when the email was sent, then your suppression would be if there are 1 Pending for Email suppress the action for 7 days.
For the Web example the outcome would be Clicks and the suppression would be if there are 1 Clinks for Web suppress the Issue for 30 days.
As for when the contact policy is used in the NBAD Framework, it's in the Customer Insights strategy within the Constraints strategy.
Behavioral Limits is the strategy that drives the precomputation of the suppressions and places them in the Action Insights dataset.
The email scenario would not work as described, because Capture Response (including Behavioral Limits - and hence the Action Insights update) is not triggered until a response is received - e.g. an email open. However, if the customer has not opened the email, it's more or less equivalent to them not having received it - since the customer has not yet read it, so starting the suppression countdown when it is read seems to make more sense.
It's understood that simply seeing the subject line could be considered to be a "soft" read, but email saturation can be controlled by creating a channel contact policy.
The web scenario should work as described since Capture Response is triggered on impression.