Closed mattmathis closed 7 months ago
Merging is not a problem, because each tag or attribute has its own column, and PMP only asserts the columns it owns, (PMP*), so it never interferes with non-PMP columns.
Note that PMP can simultaneously update multiple audiences.
The MC "segments" can do logical operations on tags or attributes, so we can construct an all member segment.
Since the initial import did not include all member types and was to the wrong audience, plan to delete the audience entirely and re import PMP into the SCVFA audience. Steps:
Please export your PMP audience and SCVFA audience. Color code PMP, add records to SCVFA table and use Excel find duplicates by highlight color. Any highlighted records left after are not already in SCVFA but I believe they are all in there except any new members since beginning of January. We can then change the export destination to the SCVFA audience. If that is your goal, why not switch the export destination now. Should we?
The PMP audience within SCVFA is just testers and people who happened to join at the right time. They should all be overwritten with fresh data. (step 3 above).
I am doing the matching in python, not excel, and have confirmed how the API interacts with existing manual data. The key is that they use different columns, and do not overwrite each other's data. MC's Tags each have their own column, so adding or deleting one does not affect any other Tags. I want to move the PMP generated data into the SCVFA Audience. I can define a MC 'segment' which matches all PMP members (since PMP labels each level differently)
Please let me move the PMPro data to the SCVFA audience- I can tightly coordinate with other potential editors, and take snapshots at the right times to robustly avoid any lost data.
Yes, please do and let me know when PMP audience is deleted so we can remove references in PMP dashboard.
Everything above is DONE.
There are now three new MailChimp segments (i.e. expressions on columns):
I manually reviewed all of item 2, and added 9 people to Susan's manually curated list. The remainders are all corner cases:
At this point MC is primed to send the Feb newsletter, using my update to Susan's manually curated list.
Still to be done (reduce to p2 for the time being):
Can we close any of the 13 issues remaining?
Test the following sequence:
(Test by manually adding a test user to MC, in the PNP audience)
Design an efficient MC update process, given the outcome of this experiment.