Alliance Bernstein - Adobe and Pardot
BACKGROUND: Internal raw data assets coming from Alliance Bernstein.
GOAL: Initial calls focused around unifying Pardot data and coming up with a Knoema match logic for Adobe ↔︎ Pardot. End Goal Still TBD
MATCHING: Relays on the “Pardot Cookie (Visitor ID)” and “prospect_id” and “ECID“. The “ECID” is the adobe ID, which can have multiple identifying events. The data also suffers from false positives on identifying/matching. This is due to Forwarding.
Forwarding is when the user who is sent the email (email creates identifying events) forwards the email on to someone who triggers the link/identifying event. This causes the tacking to be incorrect and adobe web page views data is assigned to wrong individual. Can be solved by profiling the IP address flows to identify Alliance Bernstein traffic (client forward back to Alliance Bernstein).
https://docs.google.com/spreadsheets/d/13d-n2k1xJ_DGTc0OdQnTMx4H_eVG_RXTnBe7HLE5SSs/edit#gid=0
File Type: Full File (Overwrite)
(HISTORICAL INFO)
OLD: Sample Data Provided by Alliance Bernstein.
OLD Data Objects:
Adobe.csv
AdvisorList.csv (key: AeEntityID)
AdvisorListSalesAUM.csv (key: AeEntityID)
ClientEngagement.csv ((key: AeEntityID))
RM_RC_YTD_ACTIVITY.csv (key: AeEntityID)
BuyingUnits.csv (key: AeEntityID)
“AeEntityId” attribute looks to be common key across all but “Adobe.csv”.
OLD: Sample Cut. Jira = https://knoema.jira.com/browse/CLI-42324?atlOrigin=eyJpIjoiNjRmNTI4YmIzYjBjNDAwZDljZWRjYTlmZmUwNWQxYTYiLCJwIjoiaiJ9
OLD: Notes on sample - https://docs.google.com/spreadsheets/d/1xDXbUMQO2F_a_KjPyz6_sbucvG89Y16x4JRDmctA5HM/edit#gid=478017906
Comments
@Balaji Subbaraman yes, I have instructed @Meiting Liu (Unlicensed) to fill her relevant information into this confluence page. There will be more context provided to this pipeline.
Issue we have right now is this is internal data and has no data dictionary. I agree data dictionary is required for data product ingestion, but this is raw internal data assets coming from what I would describe as a “non-technically proficient” client.
Now that I think about it, I will include that background at the top of the page.
@Balaji Subbaraman @Graham Tolson (Deactivated)
Hi Balaji and Graham, unfortunately I don’t have more information on this project either. But I have updated the notes from the most recent call (07/28) with them (old notes from 07/22 are included too):
Also, this is a Jira ticket where Emily updated the next steps expected from AB:
https://knoema.jira.com/browse/CLI-42582
Data dictionary is yet under discussion.
@Graham Tolson (Deactivated) - May I also describe a bit about what is this data all about? I have been facing challenges in understanding the nature/business value of these data. Here again, we need Data Dictionary that explains what each column means in the file; what data type it belongs to.. These are very important information we require to understand the data better during the triage stage. Thanks