Skip to content
Datauntangler

No card required. The browser demo is free to use.

What is data scrubbing, and the rest of the questions

Data scrubbing is the same job as data cleaning: finding the records in a file that are duplicated, wrongly formatted or damaged, and putting them right. Below is every question people ask before they trust a tool with a customer list.

10 questions answered on this page

0 bytes of your file leave the browser

7 terms defined in the glossary below

Clean a file Ask something else
01 Everything asked, in order

Answers, in the order people ask them

Does my file get uploaded anywhere?
No. The demo runs entirely in your browser using JavaScript on your own machine. Your file is never sent to us, never stored and never seen by anyone. You can check it yourself: open your network tab, load a file, and watch that no request goes out.
What does it actually fix?
Exact and fuzzy duplicate rows, mixed date formats normalized to ISO 8601, currency and number formats written as text, stray whitespace and invisible characters, shouted or all-lowercase names, encoding damage where Muñoz turns back into Muñoz, postcodes and IDs that lost a leading zero, and US state and country variants. It also infers what each column holds so the right rule runs on the right column.
Will it change something I did not want changed?
No. Nothing is applied until you approve it. Every proposed change is one line showing the original value, the new value and the rule that fired, and you approve or reject it row by row or in bulk. Your original file is never modified: you download a new file.
How is this different from Remove Duplicates in Excel?
Excel matches exact values only, so Acme Corp and Acme Corp. both survive, and Jose in one row and JOSE in another are treated as two different people. It also deletes rows without telling you which ones went. Datauntangler matches fuzzily across name, company and email, and shows you every row before anything happens.
How is this different from OpenRefine?
OpenRefine is a free, open-source desktop application you install and drive with its own expression language, GREL. It is a capable tool if you have time to learn it. Datauntangler opens in a browser with nothing to install, proposes the changes for you, and gives you an audit log of what you approved.
How big a file can it handle?
The browser demo reads CSV and TSV files up to 5MB or about 20,000 rows, because it runs on your own machine rather than on a server. Paid plans raise that: 25MB on Analyst, 100MB on Team, 500MB on Business, and custom sizes on Enterprise.
What counts as a row?
A row is one data line in a file you process, not counting the header. If you clean a 40,000 line contact export, that is 40,000 rows. Re-running the same file after changing a rule counts again, because the file is processed again. Rows reset every billing month.
Do you sell or train on my data?
No. We do not sell data, we do not share it with advertisers, and we do not use customer files to train any model. The browser demo never receives your file in the first place. What we handle on paid plans is described in plain language on the security page.
Can my whole team use it? Is there SSO and a DPA?
Team and Business plans add seats, shared recipes and roles. SSO through SAML or Okta, custom data retention, a DPA on request, a named support contact and invoicing are part of the Enterprise plan. The enterprise page lists what that plan includes.
What happens after I sign up?
You enter your email, we send you a six digit confirmation code, and once you enter it your address is confirmed. We will email you next steps and your account details shortly after that. There is no card required to use the browser demo, and the demo is not a trial with a countdown on it.
02 The terms, defined

Words this category uses interchangeably

Scrubbing, cleansing, cleaning, hygiene, deduplication and entity resolution overlap heavily. Here is what each one usually means when somebody says it.

Data scrubbing
Fixing or removing records that are wrong, damaged or duplicated in a file.
Data cleansing
The same job, usually said about a CRM or a database rather than about a file.
Data hygiene
Keeping it clean over time, as a habit rather than as a one-off project.
Deduplication
Specifically finding and collapsing records that describe the same thing.
Fuzzy matching
The technique that finds near duplicates which are not spelled identically.
Entity resolution
Deciding which records across sources refer to one real person or company.
Normalization
Putting values into one consistent form: dates, currencies, states, casing.
03 The question nobody asks

What happens when the engine gets it wrong

It will, on some column, on some file. The design of the whole tool assumes it.

How the process works

A wrong proposal costs one click

Rejecting it leaves the original value exactly as it was. Nothing is applied until you approve it, so a bad guess never reaches your data.

A misread column is visible

Column detection is a proposal too. When it reads a reference number as a date, you see it as a row of proposals you can reject together.

Your original file is untouched

The cleaned CSV is a new file. There is no in-place edit and no overwrite, so the worst case is that you close the tab.

Tell us which row

Send the column, the value it proposed and the value you expected to our email address. A few anonymised rows are enough.

04 Faster than reading

The answer to most of these is on screen in a second

Load a sample and watch what it proposes and, just as usefully, what it leaves alone.

1. Load a file

Your file is read on your own machine. Nothing is uploaded, so nothing can leak.

Or start from a messy sample

Status

Loading the sample file.

File: CRM contact export

Rows

--

Columns

--

Changes

--

Duplicate rows

--

Date formats

--

Blank emails

--

Load a file to see the diff

Drop a CSV or pick one of the messy samples. Every proposed change appears here as a line you approve or reject.

Related pages

Clean the file, and be able to show what you changed

Try the demo on your own CSV first. It runs in your browser, it costs nothing, and no card is required.

No card required. Your file never leaves your computer.

Clean a file