Hey Community
We utilize a system of (really two) numeric IDs in the URL string after a user clicks on an ad that represents which exact creative (d_creative=) and which advertising channel, audience, etc (sc_dcm=) the user saw/clicked on to make it to our website.
So a final URL string could look as such for two different users:
URL: http://www.example.com/path1/path2/path3?d_creative=12345&sc_dcm=678910
URL 2: http://www.example.com/path1/path2/path3?d_creative=34567&sc_dcm=78911
These IDs all then live in flat, simple (automatically updated) data lookup tables that translate the numeric values to a human readable name.
e.g. for the two creative values:
Creative ID | Creative Name
12345 | product1_cta-example-a
34567 | product2_cta-example-a
... | .... (~ 10,000 rows)
or the two Media Placement values:
Placement ID | Placement Name
678910 | facebook_audience1
78911 | youtube_audience1
... | ... (~ 8,000 rows)
These two tables get pushed into Adobe Analytics correctly (and don't really change/receive new IDs too often - read: even when a new media initiative launches, within 24 hours, the tables are automatically updated and pushed to AA via a 3rd party solution that also seems to have integrations with AAM that we don't use today), and we frequently use 'Placement Name/Creative Name contains' to build segments to analyze/understand media audience behavior better.
BUT - we have not been able to utilize the human readable names within Adobe Target to drive real-time personalization activities.
Core question: How could we use these tables to define real-time personalization/split testing?
Example:
users with creative name contains 'product1' and Placement Name contains 'audience1' -> display product 1 image on page with copy most likely relevant to audience 1
We have historically had to revert to manually inputting 100s of individual creative IDs or placement IDs into Target (as those are present in the actual URL string), which is an extremely error prone and manual process, involving multiple teams.
Any help/suggestions are appreciated. We really would like to avoid having to change our URL structure to include more parameters, or text-based parameters.
Thank you, and happy Friday!
Solved! Go to Solution.
Topics help categorize Community content and increase your ability to discover relevant content.
hi @tf-a - Adobe team has done extensive documentation on how to use Recommendations here: https://experienceleague.adobe.com/docs/target/using/recommendations/recommendations.html?lang=en
For the solution I proposed, its quite bespoke - more than happy to jump over a call and show you at work (have created a quick demo)
Regards
Rajneesh
@tf-a - an interesting challenge!
I will suggest that Recommendations can be used as a lookup table however with some limitations, including
1. multiple Adobe Target calls (which can be controlled by running this setup only when URL contains certain parameters) and
2. a flickering effect (which can be avoided by having a specific HTML module for the personalised experience, rather than changing an existing module)
Let me know if you are ok with these workarounds - I will be happy to provide more detailed solution.
Regards
Rajneesh
We already work with Minitab and are uploading that look up table within Analytics. Those values are actually evars. Would there be a way to pass the evars into mbox or is there another way for Target to access this data within Analytics.
@mattwhitehead28 - not too familiar with Minitab's capabilities. Assuming the evars you are mentioning are just the keys, you'll still need a lookup table to fetch the name-values. If you are uploading the data using Classifications then you may access it only through Adobe Target Recommendations.
Thank you so much, @Rajneesh_Gautam_ ,
Can you give us a few more pointers on where to look for documentation on your proposed solution?
Taking Minitab out of this for a second: In a nutshell, we have a relatively 'flat' .csv table of these values (they just happen to be pushed to AA as evars).
Where do you think we would need to push that table, ideally? Adobe Target (to access Recommendations), or into AAM instead? (or is there even an alternative solution that uses AAM to trigger the lookup rules?)
To your prior note (very first response):
1. We are not worried about a flickering effect (have a spinning wheel effect that covers that for the visitor).
2. multiple Target calls - while not ideal - could still be worth it, depending on cost associated with that vs. the FTE hour cost our current process creates.
We would try to limit calls by
- only triggering the rule on specific web URLs
- if a certain number of URL parameters is present
hi @tf-a - Adobe team has done extensive documentation on how to use Recommendations here: https://experienceleague.adobe.com/docs/target/using/recommendations/recommendations.html?lang=en
For the solution I proposed, its quite bespoke - more than happy to jump over a call and show you at work (have created a quick demo)
Regards
Rajneesh
Views
Likes
Replies
Views
Likes
Replies