Sigmera for Recruiters
Recruiting runs on spreadsheets that nobody designed. A longlist is a LinkedIn export stapled to an Indeed CSV stapled to a referral sheet, and the ATS wants it in a shape none of them are in. Sigmera cleans that file in a browser tab — deduplicate, normalise phone numbers, split names, fix the email column — and the candidate data never leaves your device.
The data problems you hit
- The same candidate arrives from three sources with three contact details — [email protected] from the job board, [email protected] from a referral, a work address from LinkedIn. Excel's Remove Duplicates does not catch any of it, because it only deletes rows where every value matches another row exactly.
- Applicant phone numbers come in every format a human can type: 07700 900123, +44 7700 900123, 0044 7700900123, (07700) 900-123. Workable's importer accepts only the characters 0-9, parentheses, plus, hyphen and period, and any dialler or SMS tool expects E.164.
- Job boards and CV parsers hand you a single Name column. Greenhouse's bulk import requires First name and Last name as separate mandatory fields, so the split has to happen before the upload — and middle names, double-barrelled surnames, and "Okafor, Jane" ordering all break a naive formula.
- Sourced and scraped email columns carry trailing whitespace, mailto: prefixes, two addresses in one cell, and typos like @gmial.com. Workable requires valid formatting and comma separation for multiple addresses, so a messy column fails the import outright.
- Every merge means reconciling headers. The agency sheet says Candidate Name, the ATS template says First name, and Workable explicitly tells you not to add, remove, or modify any of the template's column headings.
- Exports carry columns you should not be importing at all — date of birth, nationality, salary history, diversity-monitoring fields. GDPR's data-minimisation principle says personal data should be limited to what is necessary for the purpose, and the fix is to drop the column before the file goes anywhere.
What you can do in Sigmera
You have combined a LinkedIn Recruiter export, an Indeed CSV, and a referral sheet into one file, and you suspect a third of it is the same people twice. Deduplicating on a normalised email column collapses the repeats before they become duplicate ATS records that two different recruiters then contact.
International applicants type their number however their country writes it. Before an SMS campaign or a dialler upload, rewrite the whole column to E.164 — a plus sign, country code, and national number with no separators — so nothing bounces and phone-based duplicate matching actually works.
Your export has one Name column and Greenhouse needs two. Splitting it in a dedicated tool handles the middle names and compound surnames that a LEFT/RIGHT formula mangles, and you see the result for every row before you download.
A sourced list has stray capitals, trailing spaces, and addresses that were never valid. Trimming, lowercasing, and flagging malformed addresses first means the ATS import does not reject rows, and your sending reputation does not take the hit from inside the system.
The importer wants a Location column; your export has separate City and Country columns. Joining them gives you the template's exact shape without touching the heading row the ATS refuses to accept modified.
Recruiters work in .xlsx and importers frequently want .csv. Converting in the browser produces UTF-8 output, so accented candidate names survive instead of arriving as garbled characters in the ATS.
Sources
- 1.Art. 4 GDPR — Definitions
- 2.Art. 28 GDPR — Processor
- 3.Art. 5 GDPR — Principles relating to processing of personal data
- 4.Bulk import candidates from spreadsheet — Greenhouse Support
- 5.Importing candidate data from spreadsheet — Workable Help
- 6.Filter for unique values or remove duplicate values — Microsoft Support
Frequently asked questions
- Is it GDPR-safe to clean a candidate list in Sigmera?
- Cleaning a candidate list is processing under GDPR — Article 4(2) includes organisation, structuring, adaptation, and alteration of personal data. What changes with Sigmera is who does it. The file is read and rewritten by JavaScript running in your own browser, so no third party receives it. Sending the same file to a cloud cleaning tool would make that vendor a processor, which Article 28 says requires sufficient guarantees and a binding contract. Browser-side cleaning removes that step from the chain. It does not exempt you from GDPR generally — you are still the controller, and your lawful basis and retention obligations are unchanged.
- Why does Excel's Remove Duplicates miss the same candidate twice?
- Microsoft defines a duplicate as a row where all values in at least one row are identical to all the values in another row. If Jane appears once with a personal email and once with a work email, the rows are not identical, so nothing is removed. Deduplicating on a single identity column — and normalising that column first — is what actually catches these.
- Which columns does an ATS bulk import actually require?
- It varies, so check your ATS's own template. Greenhouse's spreadsheet import requires First name, Last name, and Job, and notes that an email is needed if you are attaching resumes. Workable specifies accepted characters for phone numbers and requires validly formatted, comma-separated emails. Both are strict about the heading row, so reshape the data underneath rather than renaming columns.
- Do I need to install anything or create an account?
- No install. The tools run in a modern browser, and the workspace keeps working offline once it has loaded because there is no server call to make. You can preview any cleanup for free and unlimited; a free account covers three downloads a month, with paid plans above that.
- What about candidate data I should not be importing at all?
- Drop those columns before the file moves. GDPR's data-minimisation principle requires personal data to be adequate, relevant, and limited to what is necessary for the purpose, so salary history or diversity-monitoring fields riding along in a general candidate import is a problem with a one-step fix. Delete the column, then run the cleanup.
- How large a candidate file can Sigmera handle?
- There is no row or file-size cap on any plan — the limit is your own machine's memory, since the processing happens locally rather than inside someone's upload quota. Large exports that make a spreadsheet app crawl generally stay responsive.