How to Scrape a Website with Apify Actors (Beginner Walkthrough)

November 10, 2026

Apify is a marketplace of prebuilt scrapers — it calls them Actors — plus a platform to run and schedule them. Need Google Maps listings, LinkedIn company pages, or Amazon reviews? Someone already built the Actor, and you run it with your inputs.

This walkthrough of Apify covers running one, exporting results, and the cost model. No JavaScript, no proxies to configure yourself.

Is Apify scraping or a coding platform?

Both, but you can use it entirely as a scraping tool without writing code. The prebuilt Actors handle the hard parts: rotating proxies, CAPTCHA handling, page rendering, and retry logic.

If you do write code, you can build your own Actor in JavaScript or Python and run it on the same infrastructure. Most users never get there.

Step 1: Create an account and find an Actor

Sign up at apify.com. You get a free monthly usage credit — enough to run a few thousand results, but not enough for a large recurring job.

Go to the Actor store and search for your target site. Read three things before running anything:

  • Last modified date. An Actor not updated in 18 months probably no longer works against a site that has changed its front end.
  • Reviews and run count. A high run count with recent reviews is the signal you want.
  • Pricing model. Some Actors charge per run, some per 1,000 results, some per compute unit. This decides your bill far more than the base platform pricing.

Step 2: Configure the input

Open the Actor and switch to the Input tab. Every Actor has a different form, but they share patterns:

  • Search terms or URLs — what to fetch.
  • Result limits — always set one. The default can be enormous, and a misconfigured run can consume a month of credits in ten minutes.
  • Proxy settings — default country matters for geo-restricted content.
  • Output format — JSON is the default and easiest to reshape later.

Start with 20 results to verify the output shape before running 2,000.

Step 3: Run it and check the dataset

Hit Start and watch the log. A working Actor streams results into the dataset tab as they arrive. If the log fills with retry messages and no results, stop it — do not let it burn credits retrying a blocked endpoint.

Check the dataset columns against what you expected. Scraped data is often messier than the sample: missing prices, inconsistent currency fields, empty images. Find that at 20 rows, not 2,000.

Step 4: Export or connect the data

The dataset tab exports to CSV, JSON, XLSX, and JSONL. There is also a direct integration to Google Sheets and a REST endpoint to pull data programmatically.

For a one-off: download CSV. For a recurring job: schedule the Actor and connect its dataset to Google Sheets, so each run appends rows to a tab. The REST endpoint https://api.apify.com/v2/datasets/{id}/items?format=json requires an API token but is how most pipelines consume results.

Step 5: Schedule it

Scheduled runs are where the cost actually accumulates. A daily run at 12:30 AM is fine for price tracking. An hourly run over large pages is not, unless the data genuinely changes hourly.

Set the task schedule conservatively and watch the first week of usage before tightening it to your real needs.

What does Apify cost in practice?

Two layers. Platform usage is billed in compute units — roughly memory-gigabyte-hours — plus data transfer, with a free monthly credit and paid tiers above it. Actor-specific pricing sits on top: some popular Actors charge a separate per-result fee that is often the dominant cost.

A realistic example: 1,000 product listings at $0.002 per result plus a few compute units lands in single-digit dollars. The same 1,000 listings from a site whose Actor charges per page can cost ten times that. Read the Actor pricing before you scale, because it varies wildly and the search results do not always make it obvious.

Where this approach fails

Highly protected sites. Marketplace Actors get blocked, and when a site adds aggressive bot detection, the Actor is usually broken until the maintainer updates it — days or weeks.

Also: sites whose terms prohibit scraping. Apify gives you tooling, not permission, and the legal exposure is yours. For personal data, GDPR and similar rules apply to what you collect, not just how.

Finally, large one-time jobs. If you only need a dataset once, the Actor may already publish a free sample dataset you can download without running anything.

Quick Summary

  • Apify is a marketplace of prebuilt scrapers called Actors, runnable without writing code.
  • Check Actor last-modified date, run count, and pricing model before running anything.
  • Always set a result limit and test with 20 rows before a full run.
  • Two cost layers apply: compute units plus per-result Actor fees, which often dominate; free tier covers a few thousand results.
  • Avoid it for heavily bot-protected sites and for any site whose terms prohibit scraping — tooling is not legal permission.
Tags: