Setting up a YouTube Signal

Creating a YouTube Signal takes about two minutes. Choose between keyword monitoring (search all of YouTube) or channel monitoring (follow specific channels), configure what to capture, and set your schedule.

Written By Kevin Lawrie

Last updated About 2 hours ago

Step 1 — Choose your Signal type

Select one of the two YouTube signal types:

Monitor for YouTube Mentions (keyword search) Discovers videos across all of YouTube that mention your keyword or phrase. Works like a saved YouTube search — results delivered automatically on your schedule.

Monitor YouTube Channels (channel monitoring) Watches specific YouTube channels for new uploads. Every video they publish is captured and processed.

Both support comment capture, notifications, and AI Agent processing.


Step 2 — Name your Signal

Click Create Social Signal in the left sidebar. Enter a descriptive name — for example, "YouTube — Competitor: Acme Corp" or "YouTube — Keyword: outbound sales tools".


Step 3 — Configure the Signal

Keyword monitoring

Setting

What it does

Options

Keyword or phrase

The search term used to discover videos

One keyword or short phrase (up to 5 words)

Include videos from the last...

How far back to search on first run

2 days, 5 days, 1 week, 2 weeks, 1 month, 2 months, 3 months

Maximum results

Max videos to collect per run

10, 20, 30, 40, 50, 100, 150, 200

💡 Tip: Keep your keyword tight and specific. Broad terms return high volume but lower relevance. Start with your brand name, a competitor name, or a specific pain point phrase your ICP uses.

Channel monitoring

Setting What it does Options

Channel URLs or handles

The channels to monitor

Add full URLs or @handles, one at a time

Sort videos by

Order for processing

Most Recent, Most Popular, Most Relevant

Include videos from the last...

Lookback period on first run

2 days, 5 days, 1 week, 2 weeks, 1 month, 2 months, 3 months

Maximum results

Max videos to collect per run

10, 20, 30, 40, 50, 100, 150, 200

You can add multiple channel URLs or @handles to a single channel monitoring Signal.

⚠️ Note: The lookback period applies to the first run only. After that, each run picks up videos published since the previous run.


Step 4 — Enable comment capture (optional)

Toggle Export Video Comments to collect top comments on each discovered video.

When enabled:

  • Comments are fetched automatically after each video is processed

  • Up to 100 comments per video are captured by default

  • AI Agents can target YouTube Comments as a data target — in addition to YouTube Videos

Credit cost: ~5 credits per ~50 comments

💡 When to enable comments: Enable comment capture whenever your use case involves understanding what people are saying in response to videos — buyer intent mining, brand sentiment, competitive reaction tracking. If you only need the video content itself (transcripts, titles, descriptions for competitive intelligence), you can leave comments off to reduce credit usage.


Step 5 — Configure notifications

Email notifications

Toggle on and enter one or more email addresses (comma or semicolon separated) to receive alerts when new videos are discovered.

Slack notifications

Toggle on and select a Slack channel. Slack must be connected in Integrations first. Notifications include video thumbnail, title, channel, engagement metrics, and a Watch on YouTube link.

Webhook

In the Export Videos section, enable webhook integration and select your webhook endpoint. Videos are sent one per HTTP call as they're discovered. See the Webhooks collection for payload schema details.


Step 6 — Assign AI Agents (optional)

Select AI Agents to process videos and/or comments automatically as they're discovered. See [Using AI Agents with YouTube] for the full guide.

Available data targets depend on your comment capture setting:

  • With comments enabled: YouTube Videos and YouTube Comments agents both available

  • Without comments: YouTube Videos agents only


Step 7 — Set the schedule

Choose how often the Signal should run: Daily, Weekly, Every 2 Weeks, or Monthly.

For competitive monitoring and brand tracking, daily is recommended. For market intelligence and trend analysis where recency is less critical, weekly works well.


Viewing results

Once the Signal runs, go to Signals → View Results to see discovered videos. Each video shows title, channel, engagement metrics, and thumbnail. If comment capture is enabled, expand any video to see its comments.

If AI Agents are assigned, their output appears in the results view alongside each video or comment — tags applied, summaries written, and any actions triggered are all visible.