Expand my Community achievements bar.

Get ready! An upgraded Experience League Community experience is coming in January.

Language Translation for Analytics Data Capture

Avatar

Level 2

Hi Community,

 

I am researching ways Adobe supports language translations for its Analytics. 

 

Context: The product team will launch a page in Japanese characters. We capture page level data via (data layer - digitalData) and engagements through  data attributes on the clickable components. Both the data layer and the data attributes are connected to the character set inputs in AEM where content authors will input Japanese characters which will have downstream impacts to our analytics

 

Are there solutions / methods we can employ to capture data in English? 

 

NOTE - Our analysts do not speak Japanese. 

Topics

Topics help categorize Community content and increase your ability to discover relevant content.

2 Replies

Avatar

Community Advisor and Adobe Champion

Hi @YohanKh1 

Short answer: no, there is no translation available in Adobe Analytics

 

This is more of a general problem when you are relying only on the visible text of a clicked link.

 

Remember when you first installed Google Chrome? It asks you whether or not to translate pages in into your set browser language.

This setting will actively rewrite the visible text on the website, leading to e.g., Arabic click text in your report suite.

 

So, I think you have two options

  1.  consider using classifications for manually uploading more context to the dimension you want to translate
    1. in this case, you can download the list of all dimensions in a given time frame, and could potentially use a translator to bring them into the language you want
    2. upload the csv to the classification tool afterwards
    3. NOTE: there is no guarantee that the final translated text will be the same when translation is applied on different source languages e.g. Japanes and Chinese, so having a translation automatism may not fully aggregate the link text dimension values 
  2. with help of your developers and content editors
    1. ask developers to create a data element on your links in your CMS (assuming they can do that), which can be filled with a e.g., English text that can be used across all websites
    2. the content editors must make sure the aggregated link text value is the same across different site language 
    3. upon link click, you must read out the new data element (when you are using activity map, you can rewrite that logic as well), to be sent as part of the link tracking call

One way or another, both options require quite some work and maybe it's worthwhile asking whether capturing every generic link click is solving any analytics question. Instead, I would try to identify which call to actions are worth capturing, and on those you can have a language independent field to aggregate meaningful interactions across different site languages.

Cheers from Switzerland!


Avatar

Community Advisor

Hi @YohanKh1 

Adobe Analytics does not provide native language translation for captured values. Whatever text is passed into the hit (page name, link text, data layer values, eVars/props) is stored exactly as sent, so Japanese characters cannot be auto-translated downstream.

That said, there are two supported and practical approaches used in real implementations -

  1. Capture language-independent values at collection time (recommended)
    Instead of relying on visible Japanese text, have AEM authors or developers populate a separate, stable field (e.g. cta_id, page_key) in the data layer or data attributes. Send this English identifier to Analytics for reporting, while Japanese text remains purely presentational.

  2. Use classifications to map Japanese to ----> English (operational workaround)
    You can classify Japanese values to English labels via the classification uploader. This works, but it’s manual, requires ongoing maintenance, and translations are not always 1:1, so aggregation can become inconsistent over time.

There is no fully automated or analyst-side translation solution in Adobe Analytics today.

For teams that don’t speak the source language, the industry best practice is to decouple analytics identifiers from visible text and track using consistent, language-neutral keys.