Update Excel data — append vs replace, and the strategic call
Re-upload the same template to refresh the data. The mechanical part is two clicks. The strategic part — append vs replace, how often, full file vs delta — is what this lesson is mostly about.
By the end of this lesson
- A refresh strategy (append vs. replace) that fits your data
- A repeatable update workflow you can hand to a teammate
You'll need
- An Excel-based app from Import Excel and CSV files
- Your refresh-template workbook with new rows added (or a full updated dataset)
Background
Once you've done a first import (covered in Import Excel and CSV files), the file you uploaded becomes the refresh template for that app. To bring in new data later, you re-upload the same workbook structure — same file name, same worksheet names, same column headers. DashboardFox replaces or appends the data inside the existing tables, and reports built on those tables keep working.
Two refresh modes, and the choice has real consequences:
- Replace — the new upload's data fully overwrites whatever's in the app. Whatever's in the file is the data, after the upload. Best when you keep the full dataset in the workbook and just add new rows over time.
- Append — the new upload's rows are added to whatever's already there. Best when you keep deltas (just-the-new-rows) in the workbook each cycle and let DashboardFox accumulate them.
One important rule: append vs replace is global to the workbook. If your app has multiple worksheets, the choice applies to all of them in a single update. You can't append to Sheet A while replacing Sheet B in the same upload. That constraint shapes the strategy below.
The strategic call: how do you plan to capture new data?
This is what to think through before your first refresh. The right answer depends on your data's nature, not DashboardFox.
Cadence. Monthly close numbers are predictable; ad-hoc operational logs aren't. Predictable cadence makes either strategy viable. Irregular updates push you toward keeping the full dataset in the file and replacing — there's less to track.
Volume. If your dataset is a few thousand rows, keeping it all in the workbook and replacing every cycle is the easiest path — easy corrections (just edit the master file), no duplicate risk, full audit trail in the file itself. If you're past 100,000 rows, the file is unwieldy; deltas + append start to make sense, but so does graduating to a database.
Correction patterns. The "keep full file, always replace" pattern wins big when you discover an error in last quarter's data. You fix it in the master workbook, re-upload, done — corrections propagate everywhere reports look. With delta + append, you'd need to upload an offset row, which is fiddly and error-prone.
Duplicate avoidance with append. If you append the same delta file twice (easy mistake), you've doubled those rows in the data. There's no automatic dedup — the file is the source of truth and DashboardFox trusts what you upload. If you append, build a discipline: keep a clear "uploaded/" archive folder, log dates, and double-check before clicking.
You're not locked in. You can use replace today, switch to append next quarter, switch back. The choice is per-upload, not per-app.
When to escalate to a database. If you find yourself building elaborate processes to keep the workbook current — multi-step exports from another system, weekly delta-extraction routines, dedicated "refresh procedure" docs for teammates — the data probably belongs in a database, not a file. The cost of curating the spreadsheet has overtaken the cost of running an actual database connection. Data connectivity overview covers your options.
Not sure which strategy fits your situation? Email [email protected] with a sentence about your data — how often it updates, how big it is, where it comes from. We'll suggest a strategy. Real human, same business day.
Do it
-
Update the template file with your new data
Open the same workbook you imported originally. The rules:
- Don't change worksheet names. Reports break.
- Don't rename existing column headers. Reports break.
- Don't reorder existing columns. Reports may break (depends on how App Builder is set up — assume it breaks).
- Adding new columns is fine. They appear as new fields in App Builder, no existing reports affected.
- Adding rows is the whole point. Add as many as you need.
- Removing rows is fine but only with replace mode — append doesn't honor removals.
For replace, your file should hold the complete current dataset. For append, your file should hold only the new rows since the last upload — nothing that's already in the app.
-
Open the app and start the update
In DashboardFox, go to Settings → Integrations → Active Integrations. Find your Excel app and click Edit. On the edit screen, click Update Data.
The update dialog opens with a file picker and the append/replace toggle.
-
Choose append or replace, then upload
Toggle Append or Replace — remember, the choice applies to every worksheet in the workbook for this upload. Pick the file. Upload.
Same as the first import, you get a preview where you can review the data and adjust types if anything looks wrong. Process. Done.
-
Verify reports still work
After every refresh — especially the first one for a new strategy — open one of the existing reports built on this app. Confirm the row count moved the way you expected (rough check), spot-check a value or two, and look for any "field not found" errors. If something's off, the most likely cause is a column rename or sheet rename in the workbook; see the pitfalls below.
If you're stuck
The most common refresh mistakes.
I appended the same file twice and now I have duplicates
Append doesn't dedupe. To recover: switch to Replace for one upload with the correct full dataset (or just the rows you wanted), then resume your appending pattern next cycle. Process: rebuild your master file with the right rows, run a Replace upload, then archive that as the new baseline.
To prevent it next time: keep an uploaded/ archive folder for delta files you've already pushed, and rename them with the upload date. Eyes-on-the-folder is the most reliable safeguard.
I renamed a column in my workbook and now my reports are broken
Reports reference the original column name; renaming breaks the link. Fix: rename the column back to its original name in the workbook, save, re-upload. (You can rename the user-facing label inside App Builder without breaking anything — that's the right place to do display-name changes.)
I renamed a worksheet and now App Builder shows it as a new table
Worksheet name = table name. Renaming creates what looks like a new table; the old one still has all the prior data, but no rows are getting added to it. Fix: rename the worksheet back to its original name and re-upload. If you actually want a different table name, that's an App Builder concern (display name override), not a worksheet rename.
I changed the structure of my workbook and now everything's a mess
Big structural changes — adding/removing/renaming sheets, reordering columns, swapping data types — can cascade through reports. Fix: restore the original structure if you can, re-upload to get back to a known-good state, and plan structural changes more deliberately. For non-trivial structural changes, consider creating a new app from scratch on the new structure rather than trying to evolve in place.
I'm looking for a "schedule refresh" option and can't find it
File imports don't auto-refresh. DashboardFox can't reach back into your filesystem or shared drive to grab the latest version of the workbook — that's a security boundary. Refreshes are always manual re-uploads.
If you want automation, you have two paths: (1) move the data to a database and let DashboardFox query it live (no refresh needed), or (2) move the data to a Google Sheet or other API-accessible source and use the API path with a scheduled fetch. See Manage API fetches.
I switched from append to replace and lost data
Replace overwrites with whatever's in the upload. If your file held only the new delta (because you'd been appending), Replace just made that delta the entire dataset. Recover by uploading the full historical dataset — usually you'll need to gather it from your archive folder of past delta files (or from the source system) and consolidate into one workbook, then Replace with that.
To prevent: when switching from append to replace, prepare the master file deliberately first. The strategy switch is per-upload, but the file content needs to match the strategy.
None of these match my situation
Email [email protected] with what you tried and what's happening. Real human, same business day.
