Trying to tie out Adobe Workspace to Downstream Databricks | Community
Skip to main content
skatofiabah
skatofiabah
Level 5
December 30, 2024
Question

Trying to tie out Adobe Workspace to Downstream Databricks

  • December 30, 2024
  • 2 replies
  • 1973 views

Hi,

 

I'm trying to tie out UVs, Visits, and Page Views from Adobe Workspace to Downstream Databricks from the adobe daily web traffic table. Here is my code. I don't know why it's not lining up and I cannot seem to find a good answer. I used the documentation to have certain filters. My page views by month start to fly off, along with visits. UVs seem to be fine though. Can anybody help with this? If not, can someone point me to a Downstream analytics consultant?

 

 

 

Thanks!

2 replies

MandyGeorge
Community Advisor and Adobe Champion
MandyGeorgeCommunity Advisor and Adobe Champion
Community Advisor and Adobe Champion
December 30, 2024

The first thing that comes to mind for me is whether or not you're using a VRS in workspace? If so, any segments that are applied to the VRS would need to be recreated in code to make the data match. Same for any bot filtering or IP exclusions that are being done to the data. 

For the page views being too high, do you have an event that fires on page views. If so, have you tried counting instances of that event? 

    skatofiabah
    skatofiabahAuthor
    Level 5
    December 30, 2024

    Hi @mandygeorge,

     

    Yes, so I think my SQL commented out code if I included that would be the mimicking the VRS I think. I commented it out to just compare to the regular report suite and I'm still getting these discrepancies. Also, what bots or IP filters would I need to exclude? My numbers downstream fall short in the more recent months, so filtering would only reduce it. Also, shouldn't Adobe's Out-of-the-Box Page Views metric upstream and downstream still align anyway? If I do do a page event instance, how do I get regex or custom code to only filter that out of event_list? I did find a report suite filter in my code that I added here (username), but I still cannot get a match.

     

    I'm close in UVs and visits overall for almost all 2024. I'm only 12 UVs less downstream and about 202 Visits short downstream. However, I am 50,000 Page Views Short in Downstream vs. Workspace. I might be missing more filters potentially? Any help or thoughts?

     

    Thanks!

     

    skatofiabah
    skatofiabahAuthor
    Level 5
    December 30, 2024

    There's an option for "exclude by IP address" in the report suite settings, I'm not sure if those automatically apply to the data feeds or not. 

    Have you tried breaking it out by week or by month to see if there is a particular time period that is off more than the others? It's possible that some of the hits could be missing or didn't come into your data lake properly. 

    How big of a variance is 50k page views for you? Depending on the size of your report suite it could be a big or small amount. For our data feeds, we have a 0.6% variance between them and workspace. You're never going to get 100% accuracy, but if you're within 1%-2%, that's generally pretty good. More than that I would look at what is happening. 

     

    As for getting the data out of the event list. You will need to know what each value in the event list means, there should be a lookup file called "event" that has a mapping. It will have data like this:

     

    This is a sample of the code that we use for pulling information from the event_list, and from the product_list. We have a CTE that unnests the two of them, and then the events are easy to pull from that. Also, the reason I use "min(date)" is because a visit can span more than one day, so I always take the day that the visit starts on.

     


    Hi @mandygeorge,

     

    Thanks for this! We saw our UVs and Visits Variance go from average 0.5% from Jan to August, to higher than 2-3% variance Sept and onward. I added this in for our VRS. The VRS just has Prop24 not equal to and eVar75 equal to filters. Any thoughts? Why it would have such a drastic change when our VRS segments and the filter itself didn't change? Our variance is essentially good for non-VRS. But with my example, if you or anybody else could help, that would be great!

     

    Thanks!

     

    Sukrity_Wadhwa
    Community Manager
    Sukrity_WadhwaCommunity Manager
    Community Manager
    March 3, 2025

    Hi @skatofiabah,

    Were you able to resolve this query with the help of the provided solutions, or do you still need further assistance? Please let us know. If any of the answers were helpful in moving you closer to a resolution, even partially, we encourage you to mark the one that helped the most as the 'Correct Reply.'
    Thank you!

    Sukrity Wadhwa
    Terms & ConditionsAccessibility statement

    Loading footer...