Processing data after receiving from data layer

Avatar

Avatar

H_L

Avatar

H_L

H_L

04-02-2021

we've set "Product ID" as evar7 to capture the product info.

The product id looks like: LAPTOP002001, PC200345, the URL looks like: https://example.com/product/laptop002001

 

Recently, an tech implementation made the URL structure changes: https://example.com/product/laptop002001-dell-xps-13

as a result, server call fires: LAPTOP002001-DELL-XPS-13 to evar7.

Then, in Adobe, we will see 2 values, they are actually same page, same product

evar 7product views
LAPTOP0020031800
LAPTOP002003-DELL-XPS-13150

 

Is there a way in Adobe Analytics process the evar and to classify to same product ID? Or it can only accept what is fired from data layer?

evar 7product views
LAPTOP0020031950

 

Thank you.

Accepted Solutions (1)

Accepted Solutions (1)

Avatar

Avatar

evolytics_brian

Avatar

evolytics_brian

evolytics_brian

04-02-2021

@H_L 

If I'm reading your question right, it sounds like you are pulling the product information from the URL (or, your data layer pulls it from the URL and that's what gets passed to Adobe). There are a couple of ways to resolve this without getting your dev team immediately involved, but the "right" approach depends on whether or not you have a tag manager (ie// Launch) in place and have the ability to make changes there.

Option 1, assuming you have access to Launch

Modify the value being passed to eVar7. This could be as simple as creating a data element where you split the value on the dashes and return the first portion. Then use that data element to populate eVar7.

For example, assume your data layer property is named "productName" and is assigned the value "laptop002001-dell-xps-13". In that scenario, the following logic would return "laptop002001", which is what I think you're looking for: productName.split("-")[0];

Option 2

If you don't have access to Launch, and you need to account for both existing data and new data, I'd suggest setting up a classification report based on eVar7. Here you can either create classification rules that use regex to do a similar parsing as above, which would account for any/all new values that come in (and can account for up to 6 months of existing data). Or, you could build a lookup table and upload as a tab-delimited file using the standard classification upload.

Answers (2)

Answers (2)

Avatar

Avatar

CurtisYoung1987

Avatar

CurtisYoung1987

CurtisYoung1987

04-02-2021

From what I have see it processes what in datalayer.

 

From the above looks like you are using the page url to set the value in the datalayer.

In implementations I have seen there would be an attribute you allow a user to set to overwrite the url reference (so if product id analytics attribute empty use url - if populated then use populated value).

 

This will allow you to keep historical data if url changes do occur.

 

A second consideration would be making use of segment builder in Workspace to build a segment that looks like:

 

PageID = LAPTOP002001

OR

PagID = LAPTOP002001-DELL-XPS-13

 

This segment can then be used for reporting as it will return the de-duplicated values for metrics in Workspace.

Avatar

Avatar

Khurshid_Alam

Employee

Avatar

Khurshid_Alam

Employee

Khurshid_Alam
Employee

04-02-2021

For historical data, the processing rules won't do much. You may take advantage of an upcoming feature in Data Repair APIs which will allow updating values in variables. It will be released later this year.

 

For ongoing data, processing rules cannot split values. Current functionality only allows concatenation. However, you may set static values for specific URLs. Not sure if that will help you much. You can also get a VISTA rule created for this.