Unlock advanced insights with the ‘Next or Previous’ function in derived fields

Transcript

Hi, this is Michele, and I’m a technical marketing engineer for the product enablement team. In this video, I’ll demonstrate the value of using the next or previous function when creating derived fields in customer journey analytics. Let’s get started. For my use case, I want to create a new dimension that concatenates up to three marketing channel values based on the next value behavior at the user level. Then I could use this new dimension to answer questions like what are the most popular combinations of marketing channels that users interact with before going in store, or which marketing channel combinations drive the highest revenue? If you’re familiar with Adobe Analytics, you might recall that cookie based JavaScript plugins or processing roles were used for this purpose. Now, the first thing I’ll do is open data views and then components. It’s worth mentioning that you need proper access to edit data views, so if you don’t see the link at the top, you’ll need to address that first. Once I’m in my data view, I’ll click the Create Derived Field button. This opens the builder where I’ll create my new derived dimension. In the left column, all available functions to choose from are displayed. I’ll locate next or previous and then drop it to the middle of the builder. I’m presented with the input fields for the role. First I’ll choose the input field under value for my use case. It’s the marketing channel. The method I’m using is next value and you’ll see previous value is available here as well. There are two options for scope person and session. Again my use case is based on the user level, so I’ll select person and the index number I’ll use for this role is relative to the base marketing channel value, which is one last. I’ll include repeats in my dimension. This setting impacts how the initial value for the first index is set. Specifically, the incidences of returning blank values. Refer to the help documentation for more details. So far I have a handle on the initial next value returned from my first role. I have two more levels remaining. I’ll add a second rule using the next or previous function below the first one. The settings are very similar to the first rule with these exceptions. The result of rule one is automatically used as the value input for rule two. This is what I want since I’ll ultimately chain the previous values together. I’ll update the index to use two, since it’s relative to the value in rule one. On to my last level. I’ll add another rule using the next or previous function below the second rule. It automatically assigns the value to the output of rule two, which is what I need. I’ll update the index value to three and use the same method scope and include repeats settings. Last. To accomplish my goal for this new dimension. I’ll add another rule. I want a concatenated value comprised of a maximum of three marketing channels. I’ll use the concatenate function for this purpose. This function allows me to combine field values together into a single derived field. Once I drop this below my last rule, I’m presented with the configuration inputs. First, I need to update the value to use the output of rule one, since that’s the first value I want to include in the concatenated string. I have the option to set the derived field only if all three values are present. The derived field would only be set if all three values exist from rules one through three. I’ll finish configuring the settings by specifying a delimiter and adding the output of rows two and three to the rule. On the right, I’ll name my field and added description. The data type for this field is automatically set. As expected, you can only use string fields with this function. Now I’ll save the new dry field. Once it’s saved, I’ll add it to my data view by dropping it to dimensions. It’s important to remember to save your changes to the data view at this point, so that the new derived dimension is available for reporting. In the workspace project, I have access to the new dimension in the metric selector, and I can use it in my report. Hopefully the next or previous function enhances your ability to answer questions about values set prior to or after an initial value, and analyze these against key events in your products. Good luck!