I have used Tiller since 2020, and am now working on “cleaning up” / refactoring my Tiller Google sheet. I created a new google sheet (with the latest “foundation template”) and have noticed some columns missing in my existing sheet.
On the “Balance History” I see I am missing the:
Balance ID
Account Status
Date Added
Source
I see “Balance ID” is documented here:
But not the other 3 columns.
Can documentation for the other 3 columns be added?
If I add these 4 columns to my existing google sheet, will it “backfill” the older data, or just enter data for “new” rows?
As far as I know it won’t backfill the older data, the balance ID and account status I have always had (or at least the last 2 years), but source seems newer; I have older data that isn’t filled out with source. That is the yodlee/plaid tracking, so not super important I don’t think. Account status is just active or not? I only have active accounts, so I am not sure what that would otherwise be.
As casilverthorn said it won’t back fill data just going forward. and unless you’re doing something specific with that sheet history doesn’t matter so long as the transaction data is complete. have you upgraded to direct fill or plaid? if not those columns aren’t as important either.
Thanks for the comments, casilverthorn.96 and ctlee !
I added those columns and confirmed it is filling them in for “new” rows, but not older ones.
I also found that my Transactions sheet (also originally created about 2020) was missing some newer fields:
* Account ID
* Categorized By
* Categorized Date
* Source
Since I am doing a “big cleanup” now (updating the categories I am using, and making some new charts n plots) I am migrating the new “Account ID” into Transactions (to clean up the “Acct #”, “Account” columns, where the names have changed over time). i.e. match rows based on “Transaction ID”, copy in the “Account ID”, then use that to update the other “Account …” columns.
When using “Direct Fills”, the “Balance History” Date column seems to now get filled with “date/time” values (i.e. values like “6/30/2026 11:31:18” instead of purely “date” values, like “6/30/2026”).
Note: I think this is “mostly harmless”, but is good to be aware of for some operations (like sorting the sheet by Date/Time/Account ID, and getting “odd” results).
And it also explains some “apparently duplicate rows” (same account / balance / date / time, but in reality the date column values differ by a couple seconds, so it isn’t quite a duplicate).
Details:
I have been “chipping away” at my “Tiller sheet cleanup”, and am now working on the “Balance History” (after finishing up with cleaning up my Transactions).
I think I now have an explanation for one of the discrepancies I was seeing while cleaning up my Balance History, and thought I would at least document it here, in case others see it also…
One thing that was puzzling me (but I think I now have it figured out) is WHY the “Balance History” “Date” column of a “newly created” Tiller sheet actually had “date/time” values in it.
i.e. while the Date column is formatted to only show M/dd/yy (like “6/30/26”), when I click on the cell the formula bar shows the actual value as something like “6/30/2026 11:31:18”.
After some experimenting, I think this is happening after “direct fill” is enabled for a sheet. The older rows (filled in by the “sidebar”) are still in the “older” format, of “6/30/2026”.
Notes:
One small surprise was that the “Date” column appears to have the highest resolution (seconds resolution), while the “Time” column is in minutes resolution. So for “highest time resolution” it appears the Date column is more accurate.
This also explained what was going on with several rows that appeared to be “duplicates” (except for different "Balance ID"s). The only “real difference” between these rows ended up being the “Date” column being off by a couple seconds (I assume small glitches when Tiller is fetching updates and gets duplicate readings occasionally).
This also explained some “weird sorting” I was seeing, where I tried to first sort by “Account ID”, then by Time, then by Date, but kept seeing some rows in the wrong spot. It turned out that while the Date column appeared to be the same for several rows, in reality they were different by a couple seconds, and that explained the sorting glitches.
@eric.lee.abcd Thank you Eric for you detailed analysis and write-up in these posts. It is helpful for both someone trying to catchup with column additions and general knowledge of these column findings. Great job!
