Get more people to use your Actor with task examples

Getting people to use your Actor isn’t easy. First they have to find it, then figure out how to configure it. Task examples help with both: each one is a saved Actor task published as its own page, with inputs pre-filled so a visitor can run it straight away.

Task examples have been around since June, but three things about them have changed. Each page now does more of the selling for you: it answers the questions buyers ask, explains how that specific task works, and is starting to show sample output. Each Actor can now have up to 10 published examples, so the ones you pick matter more than how many you publish. And you can publish them in bulk, with a script instead of one by one in Apify Console.

In short, quality beats quantity: 10 good examples do more for your Actor than 50 rushed ones. This post covers how to pick them, what the pages now do, and how to publish them at scale.

Position your Actor for what people look for

With up to 10 examples per Actor, start from what people actually search for.

I spent the last nine years building web scraping and automation workflows for clients. Most of them don’t care about web scraping. Plenty don’t even know what it is.

What they care about is more leads, more customers, and more revenue. The tool behind that doesn’t matter to them. The same goes for your Actor. It’s a tool, but the people who need it are looking for a result, so that’s what your README should speak to. Add a use case section that shows what your Actor can do.

Take my own Actor, TrustMRR Scraper. It pulls data on any startup listed on TrustMRR. Some keyword research showed that the search term most associated with TrustMRR is “monitoring”, so I published a few task examples built around monitoring use cases, then added a use case section to my README and linked the two together:

Within a month of publishing those tasks, organic traffic to my Actor page doubled, and search impressions grew 10x:

Positioning your Actors for the AI age

Task example pages are also how AI agents find and understand your Actor. Visitors who arrive from AI assistants convert to signup at 23%, more than twice the 10% rate from organic search.

Most Actors on Apify Store are web scrapers, so publishing a task example aimed at collecting large language model training data is a quick way to reposition a scraper as a tool for AI.

RAG Web Browser does this well:

Its task examples target AI-related keywords like “fetch”, “AI agents”, “Markdown”, and “RAG”.

Another angle worth one of your 10 slots is to show that your Actor can pull AI-related data. As AI goes mainstream, more sites are adding dedicated AI sections. For example:

• Medium has an AI tag, so if you have a Medium scraper, publish a task example that scrapes Medium’s AI tag.
• GitHub works the same way: you can publish a task that finds AI agent repositories.

Teach people how to configure your Actor

Task examples also show people how to use your Actor. This matters most for versatile Actors: RAG Web Browser, for instance, can either fetch a single page or run a Google search and fetch every result page. That’s hard to explain in a README alone.

AI Web Scraper is another good example:

Each page now does more of the selling

Since early October, task example pages include more than the pre-filled input:

• FAQ section. Answers to the questions people ask about the task. These are designed to help the page stand out in search results and give AI assistants something to quote when someone asks a long-tail question.
• How it works. A short explanation written for that specific task, instead of the same generic text on every page.
• A link to your Actor. Every task page links back to the Actor detail page.
• Sample output. Task pages are starting to show a sample dataset of what the task returns, so visitors know what they’ll get before they sign up. It’s being tested, so not every visitor sees it yet. The sample is AI-generated rather than taken from a real run, so no real scraped data ends up on a public page.

It’s worth checking the task examples you’ve already published to make sure they make the most of these sections.

Better pages also mean you don’t need as many of them, which fits the new publishing limit: since September 23, 2026, each Actor can have up to 10 published task examples (down from 50), and each account up to 100.

Tasks you’ve already published aren’t affected. They stay live, and you can keep editing them. The limits only apply when you publish new ones.

The limit stops AI slop from spamming Apify Store, which was hurting the whole platform, your pages included. It also works in your favor: 10 well-chosen examples rank for different searches instead of competing with each other. The Apify team will keep reviewing the limit over the coming months, so that new tasks don’t hurt search and AI visibility for everyone.

If you have a real need for more, limits can be raised for individual accounts. Soon, you’ll also see how many publications you have left directly on the task’s Publishing tab.

Publish task examples in bulk

Publishing used to happen only in Apify Console, one task at a time. It now also works through the Apify API, apify-client (2.25.0 or later), the Apify CLI (1.9.0 or later), and the Apify MCP server. You can script a whole set of examples in one sitting, or hand the job to your AI agent.

Three things to know before you script it:

1. Fill in the required fields. The task needs a title (3 to 63 characters) and a description (up to 400 characters), and its publicConfig needs inputSchemaFields and datasetView. seoTitle and seoDescription are optional and fall back to the task’s title and description. If anything is missing, the request fails with an error that lists every problem, and nothing is saved.
2. Make the Actor public first. A task can only be published if its Actor is public. If your script creates a new Actor and its examples, it needs two passes.
3. Stay within the limits. You can publish up to 10 tasks per Actor and 100 per account. If a request goes over either limit, the API returns an error, so your script can catch it.

The publishing docs cover the full setup.

Make your task examples count

Start with the task examples you already have. Check which ones match what people search for, and update the SEO title and description on the ones that don’t. If you’re starting from scratch, pick three to five use cases from your README and publish a task example for each.

Not sure what your audience is searching for? Rewatch our SEO webinar for a step-by-step walkthrough of keyword research.