Create decisions
Learn how to create decisions for decision management. A decision combines your placements and collections into a single entity, so that a decision can be made to deliver the most relevant offer to the customer.
Transcript
In this video, we’re going to learn how to create a decision in decision management capabilities. A decision combines your placements, collections, and fallback offer into a single entity so that when a delivery request is made the decision engine can decide which offer is the most relevant to the individual customer in that context. Hopping into Adobe Journey Optimizer, select offers and then select the decisions tab. This will take you to the decisions list, where you can see all your decisions. You can also sort, search, and filter them in different ways.
Selecting a row or a decision name exposes additional details and actions to the right rail.
To create a decision, select create decision.
Next give your decision a name, and if you want a start and end date.
Next we need to choose the placements for our decision. In the placement box, use the dropdown menu to select one of your existing placements. You can also search for placements by name, or filter them by channel and or content type. Select a placement to see additional metadata about the placement.
We can now turn to evaluation criteria.
This determines which offers will be shown or delivered in the placement. At least one set of criteria must be configured with every placement. To begin configuring evaluation criteria select add. Here you can select which collection you want to use for your chosen placement. Remember, the collection represents the set of personalized offers that decision management will decide between. Selecting a collection will show you all of the offers in the collection using that placement. If we need to make updates to the collection we can select open offer collections, to open the collection in a new browser tab. In that tab, we can add the collection as needed. While a decision is running, we can update the offers it’s delivering simply by updating the collection. Let’s go back to the decision. Next let’s select the collection we wish to add followed by selecting the add button. Next let’s choose an eligibility constraint that will restrict the selection of offers for this placement. Our options are none, segments, or decision rule. Choose none if you don’t want to apply any eligibility constraints. Choose segments if you want to restrict the selection of offers to members of one or more Adobe Experience Platforms segments. Or in other words a specific list of profiles. You can add one or more segments, from the left pane and combine them using the and or logical operators. In this example, I’m adding the gold customer’s constraint which represents all profiles that have the gold tier attribute.
I’m also going to use the and operator and combine it with the California customer’s constraint which represents all profiles that live in the state of California. Choose decision rule if you want to execute on demand, a rule against all individual Adobe Experience Platform profiles every time a decision is made. In this example, I’m selecting the loyalty tier gold rule which represents all profiles that have the gold tier attribute.
Note the differences in how decisions use segments versus decision rules. When a decision process runs using segments, the process will run against pre-computed Experience Platform data. That is a list of all your profiles, the events associated with them, and attribute data. Segments are generally all encompassing and your decision data will feed back into platform as a bonus. When a decision process runs using decision rules the process will run against experience platforms profile data store at the run time. In other words, rules have access to offer attributes since we have the user ID at the time of the request and rules allow you to use contextual data. Contextual data refers to time sensitive offers that can quickly change, such as an offer based on the current weather outside. However, because rules are evaluated on the edge to allow for time sensitive offers, rules cannot take events into account as events do not live on the edge profile. Depending on the offer you wish to serve choose either a decision rule or a segment.
Next let’s choose a ranking method. When our chosen eligibility constraints result in multiple eligible offers for our placement a ranking method calculates which among these offers should be delivered first to the given profiles. Our options are offer priority, or formula. Choose offer priority if you want to use offer priority scores to calculate which offer will be delivered first. Recall that offer priority is a numeric score that is defined during the process to create or edit personalized offers. An offer with a higher score, will be delivered before an offer with a lower score. Choose formula if you want to use a specific ranking formula to calculate which offers will be delivered first. Ranking formulas can make calculations based on profile attributes, offer attributes, and context data. For example, we could select the ranking formula that will boost offers when the weather is hot.
Select add to define more criteria for the same placement. When you add multiple criteria they will be evaluated in sequential order. In this example, new gear will be evaluated first, then Luma collection will be evaluated second. If you want to change the sequence, you can drag and drop the separate criteria and reorder them.
You can also evaluate several criteria at the same time. To do so, drag and drop separate criteria on top of one another. In this example, new gear and Luma collection, now have the same rank, and thus will be evaluated at the same time.
To add another placement for your offers as part of this decision, use the new scope button.
Follow the same procedure for all your placements. Remember that an offer can also include multiple placements, so depending on your setup it might make sense to use the same collection for each placement. Select next to proceed, to the add fallback offer step.
The fallback offer is the offer that gets delivered if a customer doesn’t qualify for any of our personalized offers in the collection. We can only the choose one fallback offer for our decision. So we should build one, that includes all of the placements being used. By default the screen will only show us fallback offers that use all of the placements chosen for our decision. Choose a fallback offer and select next to continue.
After reviewing our decision, we’re ready to select finish and save. At this point, we can decide if we want to save it as a draft or activate it as well. If we save it as a draft, we can activate it from the activities list at any time by selecting the activity icon located in the right navigation.
Once you’ve created a decision, it will be assigned an offer decision ID. The combination of the placement and offer decision ID makes a decision scope which gets used in the API request to retrieve the correct offer for the customer. These decision scopes can also be copied from the decision page in several formats. Note that there’s a separate decision scope for each placement.
You should now be able to create decisions, and have a sense of how they’re used with placements, collections, and fallback offers. Thanks for watching. -
To learn more on how to create decisions with using the Batch Decisioning API, see the product guide.
recommendation-more-help
7e382214-bd30-4de2-bc8b-f6f6e7182305