Coming full circle with Marketing Cloud & Einstein Analytics
Marketing departments are no different than sales or finance, they too want insight. With all the data marketing is generating it becomes that much more interesting to start having a tool that can give you good insight. Salesforce have two marketing automation tools – Pardot and Marketing Cloud. Pardot already has an app that leveraging Einstein Analytics, but Marketing Cloud doesn’t, however, we do have a default connector we can use. Therefore, I decided to explore what we could do with Marketing Cloud and Einstein Analytics, which ended up being 5 blogs.
- Marketing Smart with Einstein Analytics – part 1: Explores setting up the Marketing Cloud connector in Einstein Analytics.
- Marketing Smart with Einstein Analytics – part 2: Explores how to surface tracking data for the connector for low data volume.
- Marketing Smart with Einstein Analytics – part 3: Explores how to create a dataset with the Marketing Cloud data.
- Marketing Smart with Einstein Analytics – part 4: Explores how to surface tracking data for the connector for large data volume.
- Marketing Smart with Einstein Analytics – part 5: Explores how to pass data back into Marketing Cloud with bulk actions.
December 2019 update: with the new authentication in Marketing Cloud there is a new type of installed package as well as connector to use, which is covered in the blog A new OAuth 2.0 Marketing Cloud connector in Einstein Analytics.
I hope this blog series has inspired you how to get Marketing Cloud and Einstein Analytics connected.
9 thoughts on “Coming full circle with Marketing Cloud & Einstein Analytics”
Hey Rikke, wouldn’t Datorama be a better tool then EA for marketers when they are bringing in more then just MC data? EA doesn’t seem scalable for all of the different marketing datasources..
It always comes down to use-case and tools available.
What are the use cases of using both EA&D and Datorama?
Where do they compliment each other?
I don’t know enough about Datorama to give a qualified answer. But to my knowledge Datorama really strong on connecting different marketing tools and providing a out of the box rapporting to those. EA is fully customizable but you don’t have all the connectors to marketing tools. Also I guess EA is better embedded in CRM.
Thanks for the immediate answer.
I have reached out to Gert and Oskar from SFDC 🙂
A little mind teaser on a Friday.
We are on a new project and have stumbled into an exciting issue.
We aim to add a sum or total of a column into the dataflow leaving us with a new column all containing the same value (the total of the other column). Is this possible – haven’t found anyone mentioning they have achieved this.
The idea is to use flow, not recipe.
Check out compute relative. But if it’s the sum of a whole column isn’t it just enough to aggregate on the dashboard?
HI – thing is we tried the compute relative. But I can’t figure out how to get the total of the column, even if I flag each column and set it as a partition.
The total is needed for further calculations in the dataflow, due to the issue with making a rolling total on a column based on a formula in the compare table.
hope it makes sense 🙂
This may be a bit late, and but hopefully helpful nevertheless.
I understand you want a groupby aggregation that then gets broadcast to all rows with the same groupby key in the original dataset. In Pandas, we would use .groupby(key)[‘field’].transform(sum) or something like that, but that’s not available in a dataflow. (There is a recipe way to do this as well, but I think you’re looking for a dataflow.)
The fundamental idea I’m using here is the following: Suppose that table A has columns X, Y, Z, where X is a column of identical entries “dummy”, Y is some ID that you’re interested in, and Z is a numeric column that you want to sum.
A augment A
Left id = X
Relationship = s
Right id = X
Right fields = Z
Method = “Lookup Multivalue”
what actually happens is that the augment acts as an aggregator. You’ll get columns X, Y, Z, s.Z, where s.Z will be the sum of all of the entries in Z.
If you don’t have a column like X in your data, then create one with a computeExpression node: name = X, formula = “dummy”
Incidentally, if you wanted to know the total of Z across the ids in Y, the same idea would work, replacing the id fields with Y in lieu of X.
In the end, a groupby aggregation-type node would be a good future-state, but look at recipes in Winter ’21 to get really stoked about that.