What is ETL?
Quick answer
ETL stands for extract, transform, load. You extract data from a source, such as an inbox or a CRM, transform it into the shape you need, then load it into a destination, such as a spreadsheet or database. It is the classic pattern behind data pipelines and many office automations.
Last updated
Take the free module →Module 13 of the free AI for Accounting & Bookkeeping Firms course, about 14 minutes. No account needed.
Updated · By Robert Breen
Why it matters for a small business
The term comes from data warehousing, where IT teams move data from business systems into one place for reporting. But the pattern is everywhere in office work. Pulling invoice totals from PDFs into a bills sheet is ETL. So is copying the sales pipeline into the Monday report. Naming the three stages makes it easier to see where a process breaks: getting the data, fixing it, or putting it away.
A related term is ELT, where you load raw data first and transform it inside the destination, which modern cloud data warehouses make practical. For most small teams the order matters less than the habit: keep the transform step visible and testable, because that is where wrong dates, swapped decimals and duplicates are caught or created.
In a real lesson: Invoices from Your Inbox to Google Sheets, with Approval
Invoices from Your Inbox to Google Sheets, with Approval is a small ETL pipeline in n8n. Extract: the Gmail Trigger catches invoice emails and downloads the PDF, and Extract from File turns it into text. Transform: the Information Extractor pulls the vendor, invoice number, dates and total, with dates written year first and the total as a plain number. Load: Append row in sheet writes it to Invoices 2026. The lesson then adds an If node so anything over $1,000 goes to Lena for approval.
The same three stages show up in tools that never say ETL. Docparser extracts fields from PDFs and loads them into a CSV, and Coefficient pulls 12 deals from a source sheet into the Weekly Sales Report on a daily refresh at 9:00 AM, with an optional filter on which rows come in.
Try this lesson free or read the step-by-step guide.
Common confusions
ETL vs data extraction
Data extraction is only the first E, pulling information out. ETL also covers reshaping it and saving it somewhere.
ETL vs integration
An integration connects two apps. ETL describes the job of moving and reshaping data, whatever tool does it.
Tips
- Write down the source, the transform rules and the destination before you build anything.
- Test the transform step with a sample record and check every field in the destination.
- Add a duplicate check before loading, so a resent invoice isn't logged twice.
Related terms
More Data, Sheets and Gmail terms
Where you use it: free lessons
- Invoices from Your Inbox to Google Sheets, with Approval (n8n, 14 min)
- Docparser: Pull Invoice Data from PDFs into a Spreadsheet (Docparser, 10 min)
- Coefficient: Pull Live Pipeline Data into Google Sheets and Auto-Refresh It Daily (Coefficient, 12 min)
Related topics
- Business intelligence (BI) (Business intelligence glossary): Business intelligence (BI) is the set of tools and habits a company uses to turn its everyday data (orders, invoices, tickets…
- Data source (Business intelligence glossary): A data source is the connection a BI tool reads from: a spreadsheet, a CSV export, a database table or a cloud app such as a…
- Data warehouse (Business intelligence glossary): A data warehouse is a central database designed for reporting and analysis. It holds cleaned, combined copies of data from many…
Frequently asked questions
- Do I need a data engineer to do ETL?
- Not for office-sized jobs. Tools like n8n, Docparser and Coefficient handle the extract, transform and load steps with clicks.
- What is the difference between ETL and ELT?
- ETL reshapes data before saving it. ELT saves the raw data first and reshapes it inside the destination system.
- How often should an ETL job run?
- As often as people need fresh data: every new email for invoices, or once a day for a sales report.