Dotwave Applying recipes to new files
A saved recipe is only valuable if applying it is effortless. In Dotwave it is: upload next month's file, pick the recipe, and every cleaning step you carefully worked out the first time runs again automatically. This article walks through the one-click flow, explains how Dotwave copes when column names drift between files, and covers what happens when a column in the recipe can't be found in the new data.
One-click re-cleaning
Applying a recipe to a new file follows a short, predictable path:
Add this period's file to Dotwave as a new dataset, just as you would any other upload.
From the Recipes menu, choose the recipe you built for this report — for example Monthly Sales Cleaning.
Dotwave begins replaying the stored steps against your new dataset.
Before running the steps, Dotwave matches each column the recipe expects to a column in your new file using fuzzy matching (explained below).
Once the mapping is confirmed, every operation in the recipe executes in its original order — no manual clicking through each step.
Check the cleaned result, then export it as CSV or Excel, or send it straight to a dashboard.
Fuzzy column mapping
Files exported from real systems rarely keep identical headers month to month. One export calls a column Sales Amount; the next calls it sales_amount. If a recipe demanded an exact header match, a single renamed column would break the whole run. Instead, Dotwave uses fuzzy matching to line up the columns the recipe expects with the columns actually present in your new file, tolerating differences in casing, spacing, punctuation, and minor wording.
The mapping is not hidden from you. Before the steps run, Dotwave shows you which recipe column it has matched to which file column, and you can review and override any match it got wrong. This keeps you in control: the automation handles the obvious cases, and you make the final call on anything ambiguous.
Always glance at the proposed column mapping before you confirm. Fuzzy matching is accurate for typical header drift, but you are the one who knows whether "Total" means revenue or unit count in this particular file.
When a column doesn't match
Sometimes a column the recipe needs simply isn't in the new file — it was dropped upstream, renamed beyond recognition, or the export changed shape. When Dotwave can't confidently map a column, it does not guess and quietly corrupt your data. Instead, that step is skipped and flagged so you can see exactly what was left out.
You then have two ways forward. If the column is present under a name the matcher missed, fix the mapping to point the step at the right column and let it run. If the column genuinely no longer exists and the step is no longer relevant, you can skip the step permanently so the flag stops appearing on future runs. Either way, nothing happens silently — every unmatched step is surfaced for your decision.
Recipes save hours on recurring client work. Build the recipe once — on a clean, well-understood file — then trust it every month.