DataMesh query
Hi - I just started using SmartSheet and am really excited by what it has to offer.
I started to explore DataMesh and have some feedback/queries on it;
Queries - I created a config with ~100 columns - it took me a very long time to do the mapping. Now that it is completed I realized that it links to a specific source sheet - and unless you copy & paste data into that same source sheet (which when dealing with large amounts of data is painful to do), you need to create a new Config linked to the new Source sheet. Given it took me a lot of effort to create the initial Config it would be great to have the ability to update the source to a different sheet, or be able to clone the existing Config and make slight edits without having to recreate everything. My new sheets have exactly the same column layout as the original sheets, but will just have new data in them as they get run daily.
What's the best workaround for this?
Feedback - it's time consuming selecting the source columns you want, then having to reselect them again in the destination section. Please consider improving the UI here to make it easier on us as users. Also when we have long column names the UI cuts them off making it hard to tell if you have accidentally included a duplicate. There is also no easy way to re-order or add/remove a column selection without impacting every other selection you made further down. It would be great to provide more functionality around this in case you want to re-order things after inputting 100 odd columns.
Thanks!
Best Answer
-
Hi @.Leon.
While this is possible, whether or not this is the best way to proceed really depends on your end goal, what your data is, how you have your sheets set up, and your specific use-case.
I wouldn't advise building 100 cross-sheet formulas in all your columns, especially as there is a 100 cross-sheet reference limit (see here). This is actually where DataMesh would be best, to do the work of the formulas and matching content instead. However I may have misunderstood; will your Sheet 2 (Summary Sheet) only have a few columns, not all 100? If you only want to bring through specific relevant columns (say 10 - 15) then yes, this process could work.
I would recommend setting up a Pro Desk session as a better way to discuss your options in regards to DataMesh, since a screen share will be easier to show your specific scenario and discuss ways to reach your end goal.
Cheers,
Genevieve
Answers
-
Hi @.Leon.
There currently isn't a way to clone a DataMesh configuration, as you've found. You will need to re-create the config with a new sheet, identifying each column mapping.
When the Source Sheet is an entirely new sheet, even if it has the exact same column names and types, the Data Mesh configuration sees each of these columns as completely new columns and has no way to auto-map them based on a previous configuration.
Thank you for detailing all of your suggestions, it's really good to hear about your use-case and feedback. If you have a moment, it would benefit our Product team if they could hear from you as well. Please fill out this form here with your feedback.
Thanks,
Genevieve
-
Thanks Genevieve - if I copy & paste over the existing source sheet to update it with new content - then re-run the same DataMesh config - will that work / get around this?
-
Hi @.Leon.
If I'm understanding you correctly, then yes. You can update/add content within the same source sheet and then the Data Mesh configuration won't need to be re-created, since it's already synced to the right sheet and columns.
However before you do this, your account is eligible to book a Pro Desk session. You may want to book a session under Connectors & Premium Apps so you can discuss DataMesh over screen share and identify any possible issues before running the workflow.
Cheers,
Genevieve
-
Thanks - I realized an alternate & simpler approach maybe to have 3 sheets;
i) a reference sheet for DataMesh updates that just has the current list of unique identifiers I need which then triggers updates in my Summary sheet by adding additional 'new' unique identifiers as they populate into a Summary sheet. So this just needs 1 column.
ii) a Summary sheet that is populated with all the unique identifiers currently in use - as new ones appear in the DataMesh reference sheet - they will be added to the Summary sheet.
iii) a Source data table that has all my source information across 100+ columns, with the unique identifier listed in a column as well.
In this manner my Summary table would then do VLOOKUPS for the necessary tables in the Source sheet, and I would add new unique identifiers to the Summary sheet via the DataMesh reference sheet.
I would only need to ingest 1 column into the DataMesh reference sheet making it super easy to run the DataMesh config each time. And I would just delete/overwrite the Source data table myself each time I updated the DataMesh reference sheet - this removes the burden of having to redo the DataMesh config every time.
Does that sound feasible or in line with what others do to get around this?
-
Hi @.Leon.
While this is possible, whether or not this is the best way to proceed really depends on your end goal, what your data is, how you have your sheets set up, and your specific use-case.
I wouldn't advise building 100 cross-sheet formulas in all your columns, especially as there is a 100 cross-sheet reference limit (see here). This is actually where DataMesh would be best, to do the work of the formulas and matching content instead. However I may have misunderstood; will your Sheet 2 (Summary Sheet) only have a few columns, not all 100? If you only want to bring through specific relevant columns (say 10 - 15) then yes, this process could work.
I would recommend setting up a Pro Desk session as a better way to discuss your options in regards to DataMesh, since a screen share will be easier to show your specific scenario and discuss ways to reach your end goal.
Cheers,
Genevieve
-
Perfect - thanks!
I doubt ill have that many columns in use, but it just seems easier than redoing a DataMesh config every time.
I do have some Pro Desk credits so will look to book a session with them when timing permits.
L
-
I have used a report as the source for a Datamesh - you can alter the report's source sheet to point to the latest version of the data.
