10 output formats and export behaviors every Minexa.ai user should understand

Most people focus on getting data out of a website. Fewer think carefully about what that data looks like once it arrives. Here is what you actually need to know about how Minexa.ai structures and delivers your extracted data.

1. Excel is the default, and it is already structured

When you run a scraping job in Minexa.ai, the default export format is Excel (.xlsx). The file is not a raw dump. Each result gets its own row, and each data point gets its own column. If you scraped 300 product listings, you get 300 rows with columns like name, price, rating, and availability already separated. No manual splitting or reformatting required.

2. JSON preserves relationships that flat formats cannot

Some pages contain nested data: a product with multiple reviews, a listing with several images, a job posting with a list of required skills. Excel flattens everything into rows and columns, which can distort these relationships. The JSON export keeps the nested structure intact, making it the better choice when your downstream process needs to read hierarchical data accurately.

3. Google Sheets export is a destination, not a snapshot

Exporting to Google Sheets sends your data directly into a sheet you control. If you run the same scraper again later, the new data can populate the same sheet. This makes it practical for ongoing monitoring tasks where you want results to land in one place without manually moving files each time.

4. Image links are columns, not downloaded files

When Minexa.ai detects images on a page, it captures the image URL as a data point in its own column. You get a direct link to each image, not the image file itself. This is intentional: it keeps exports lightweight and lets you decide whether to download the files separately, display them in a tool, or pass the URLs to another system.

5. Hidden attributes are surfaced automatically

Web pages often contain data that is not visible to someone reading the page normally. Product IDs, internal category codes, tracking attributes, and metadata values are embedded in the page code but never displayed. Minexa.ai detects and extracts these alongside the visible fields, so your export can include data points that you would never find by reading the page yourself.

Ready to see what your data looks like? Install the Minexa.ai Chrome extension and run your first export in minutes: Get the extension

6. Column names are generated automatically

You do not name your columns manually. Minexa.ai generates column labels based on what it detects in the page structure. Labels like "price_whole", "listing_title", or "review_count" come from the pattern recognition process. These names are consistent across all rows in a given export, but they are not guaranteed to match labels from a different site or a retrained scraper on the same site.

7. Two-layer extraction produces one unified output

When you choose to scrape both the list page and the individual detail pages behind each result, the output is a single combined dataset. You do not get two separate files to merge manually. Each row in the export contains both the data from the list and the data pulled from that result's detail page, joined together automatically.

8. Missing fields return empty, never invented

If a specific page in a batch does not contain a field that other pages do, that cell in the export is empty. Minexa.ai does not fill gaps with estimated or inferred values. This matters for data quality: an empty cell tells you the data was not there, which is accurate. A filled cell with a wrong value would be harder to catch and more damaging to rely on.

9. Retraining can shift column names

If a website redesigns its layout and you retrain your scraper, the new column names may differ slightly from the original ones. A field previously labelled "price_whole" might come back as "price_full" after retraining. If you have a spreadsheet formula, a connected tool, or any process that references column names by their exact label, it is worth reviewing those references after any retraining step.

10. The output reflects the page exactly as it is

Minexa.ai does not reformat, reorder, or interpret the data it extracts. If a price appears as "$1,299.00" on the page, that is what appears in your export. If a date is written as "June 2026", it stays in that format. There is no automatic type conversion, currency normalization, or text cleanup applied. What you see on the page is what you get in the file, which means your own formatting rules apply from that point forward.

Start extracting structured data today. The Minexa.ai Chrome extension handles detection, pagination, and export automatically: Install Minexa.ai